kidney stone patient case study

• Case Report
• Open access
• Published: 10 July 2009

Acute onset of renal colic from bilateral ureterolithiasis: a case report

• Eduardo de Paula Miranda 1 ,
• Diego Costa Almeida 1 ,
• Gustavo Pinto Ribeiro 2 &
• Ariel Gustavo Scafuri 1  

Cases Journal volume  2 , Article number:  6354 ( 2009 ) Cite this article

8901 Accesses

Metrics details

We report a case of a 32-year-old man, who presented to the emergency department with severe abdominal pain, with radiation to his back. An ultrasound examination revealed mild hydronephrosis bilaterally. A non-enhanced computer tomography was then performed and showed a 9 mm hyperdense image in the left ureter topography along together with an 8-mm hyperdense image in the right ureter topography, allowing us to establish the diagnosis of bilateral ureterolithiasis. The patient was taken to the operating room in order to perform ureteroscopy for endoscopic removal of the stones.

Introduction

Urolithiasis is a problem, which affects the humanity for centuries and is relatively common, with reported incidences of up to 12% of the world population during their lifetime [ 1 ]. Various studies estimate that 2-3% of all individuals present annually with either a sign or symptom related to urinary tract obstruction secondary to calculus impaction.

People aged from 20 to 30 are believed to have the highest incidence, especially men, who are affected three times more than women. Family history of urolithiasis is also an important risk factor, once up to 55% of individuals with recurrent stones report cases in the family [ 2 ].

The most likely mechanisms include: (i) the possible presence or abundance of substances that promote crystal and stone formation; (ii) a possible relative lack of substances to inhibit crystal formation; and (iii) a possible excessive excretion or concentration of salts in the urine, which leads to supersaturation of the crystallizing salt. Calcium stones account for 75-85% of urinary calculi. Approximately one half of calcium stones are composed of a mixture of calcium oxalate and calcium phosphate [ 3 ].

Case presentation

A 32-year-old white man, who works as a construction worker, presented to the emergency department with severe and debilitating abdominal pain. The patient described the pain as colicky and diffuse throughout the anterior abdominal wall, with radiation to his back. The pain was exacerbated by walking and relieved with rest. He referred vomiting and low fever (37.8°C), but he denied any chills or night sweats. He had no previous history of urinary calculi, but he affirmed that his brother had a kidney stone, which passed spontaneously.

The patient had no known chronic medical conditions and was currently not taking any medications. He had no previous history of urinary calculi, but he affirmed that his brother had a kidney stone, which passed spontaneously. He denied alcohol, tobacco or any intravenous drug abuse.

On physical examination was 1.74 meters tall and weighed 69 kilos. The patient appeared in distress, which improved after parenteral analgesia. He was afebrile and his abdomen was diffusely tender to palpation.

Blood urea nitrogen and creatinine were within limits of normality; the rest of laboratorial analysis was unremarkable, except for mild leucocytosis of 12,000/μL, microhematuria, pyuria (10 leucocytes/field) and presence of crystals at urinalysis. An ultrasonography was performed and revealed mild hydronephrosis bilaterally. A nonenhanced computer tomography (CT) was then performed and showed a 9-mm hyperdense image in the left ureter topography along together with a 8-mm hyperdense image in the right ureter topography (Figures 1 and 2 ).The diagnosis of bilateral ureterolithiasis was then established.

Abdominal CT revealing a 9-mm hyperdense image in the left ureter topography .

CT showing 8-mm hyperdense image in the right ureter topography .

The patient was taken to the operating room in order to perform ureteroscopy (URS) for removal of the stones. The procedure was uneventful and the patient was left with a double J stent at both sides to be removed at follow-up. These measures have provided immediate relief of the symptoms and 36 hours later the patient was discharged. The stone fragments were sent to analysis and the patient is currently attending a nephrologist for clinical control of idiopathic hypercalciuria.

Even though urolithiasis is a common affection, an acute onset of renal colic after bilateral ureterolithiasis is rather uncommon, with no similar reports in the literature.

We believe that the number of such cases is actually underestimated, probably due to spontaneous passage of the calculi before imaging tests. Besides, many other abdominal conditions may present in a similar way, leaving some cases undiagnosed.

Acute ureteral obstruction by stone usually causes severe colicky flank pain that can radiate throughout the groin, testicles, back, and periumbilical region. As the anterior abdominal pain dominated the clinical presentation of our patient, he was thought to have an acute abdominal condition, which led us to perform an abdominal ultrasonography priory to CT.

With a sensitivity of 94-97% and a specificity of 96-100%, nonenhanced helical CT is the most sensitive exam for the detection, localization, and characterization of urinary calcifications; therefore, it is considered the gold standard for approaching urinary stones. Intravenous Urography (IVU) takes more time, requires contrast and provides no additional clinically important information [ 4 ]. Thus, in places where CT is available, IVU should not be performed.

Urgent intervention by means of either percutaneous nephrostomy or ureteral stenting is indicated in a patient with an obstructed, infected upper urinary tract, rapid renal deterioration, intractable pain or vomiting, anuria, or high-grade obstruction of a solitary or transplanted kidney [ 2 ].

Distal ureteral calculi (<5 mm) usually pass the ureter spontaneously. Ureteroscopic lithotripsy of distal ureteral calculi shows high stone-free rates with a low complication rate (4%) and is equal to extracorporeal shock wave lithotripsy (ESWL), while ESWL is the primary choice for proximal ureteral stones [ 5 ]. Though the selection of these two options depends on equipments available and the expertise of the operator, URS is recommended by many authors as the optimal treatment for distal ureteral calculi [ 6 ].

The complication rate of URS is 9-11% and usually consist of avulsion of the ureteral urothelium, ureteral perforation, stricture (<1%), impaction of the instrument in the ureter with consequent ureteral laceration, extravasation of stones, and bleeding in the urogenital tract, but are minimal in experienced hands [ 5 ]. There is no evidence that bilateral approach during URS increases complication rates.

As our patient presented with 8 and 9 mm distal ureteral stone bilaterally, which is unlikely to pass spontaneously, he should be offered shock-wave lithotripsy or URS. However, our experience has led us to believe that when both ureters are affected, URS should be the procedure of choice, despite the lack of evidence in the literature.

Written informed consent was obtained from the patient for publication of this case report and accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal.

Abbreviations

computer tomography

extracorporeal shock wave lithotripsy

Intravenous Urography.

Sierakowski R, Finlayson B, Landes RR, Finlayson CD, Sierakowski N: The frequency of urolithiasis in hospital discharge diagnoses in the United States. Invest Urol. 1978, 15: 438-441.

CAS   PubMed   Google Scholar  

Teichman JMH: Acute Renal Colic from Ureteral Calculus. N Engl J Med. 2004, 350: 684-693. 10.1056/NEJMcp030813.

Article   CAS   PubMed   Google Scholar  

Finlayson B: Physicochemical aspects of urolithiasis. Kidney Int. 1978, 13: 344-360. 10.1038/ki.1978.53.

Smith RC, Rosenfield AT, Choe KA, Essenmacher KR, Verga M, Glickman MG, et al: Acute flank pain: Comparison of non-contrast-enhanced CT and intravenous urogram. Radiology. 1995, 166: 97.

Google Scholar  

Hofmann R: Ureteroscopy (URS) for ureteric calculi. Urologe A. 2006, 45: w637-w646. 10.1007/s00120-006-1035-5.

Zeng GQ, Zhong WD, Cai YB, Dai QS, Hu JB, Wei HA: Extracorporeal shock-wave versus pneumatic ureteroscopic lithotripsy in treatment of lower ureteral calculi. Asian J Androl. 2002, 4: 303-305.

PubMed   Google Scholar  

Download references

Author information

Authors and affiliations.

Federal University of Ceará, Faculty of Medicine, Rua Aleandre Baraúna 949, 60430-160, Fortaleza/CE, Brazil

Eduardo de Paula Miranda, Diego Costa Almeida & Ariel Gustavo Scafuri

Federal University of Vale do São Francisco, Faculty of Medicine, Av. José de Sá Maniçoba, S/N - CEP, 56304-917, Petrolina/PE, Brazil

Gustavo Pinto Ribeiro

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Eduardo de Paula Miranda .

Additional information

Competing interests.

The authors declare that they have no competing interests.

Authors' contributions

EPM conducted the patient, participated as the first assistant during the surgical procedure and was a major contributor in writing the manuscript. DCA participated as second assistant during the surgical procedure and was responsible for collecting data and consent from the patient. GPR was responsible for the literature review and was a major contributor to writing the manuscript. AGS was the intellectual menthor, the surgeon who led the team and the reviser of the manuscript and all collected material. All authors read and approved the final manuscript.

Rights and permissions

Open Access This article is published under license to BioMed Central Ltd. This is an Open Access article is distributed under the terms of the Creative Commons Attribution License ( https://creativecommons.org/licenses/by/2.0 ), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Reprints and permissions

About this article

Cite this article.

de Paula Miranda, E., Almeida, D.C., Ribeiro, G.P. et al. Acute onset of renal colic from bilateral ureterolithiasis: a case report. Cases Journal 2 , 6354 (2009). https://doi.org/10.4076/1757-1626-2-6354

Download citation

Received : 03 March 2009

Accepted : 29 June 2009

Published : 10 July 2009

DOI : https://doi.org/10.4076/1757-1626-2-6354

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Shock Wave Lithotripsy
• Calcium Oxalate
• Extracorporeal Shock Wave Lithotripsy
• Ureteral Stone
• Renal Colic

Cases Journal

ISSN: 1757-1626

• Sign Up Free

• Stone disease

Analysis of Risk Factors for Postoperative Recurrence in Elderly Patients with Kidney Stones: A Case-Control Study.

Renal calculi are solid crystals that form in the kidneys and cause severe pain and discomfort. This study aims to investigate risk factors for postoperative recurrence of renal calculi in elderly patients and provide background knowledge on the prevalence and management of renal calculi in this demographic.

The clinical data of 123 elderly patients with renal calculi were included from 1 June 2021 to 1 June 2023 for their 6-month follow-up study. The patients were divided into recurrence group and non-recurrence group according to whether they had recurrence after surgery. The general sociological characteristics and disease-related characteristics of the two groups were counted. Logistic regression equation was used to calculate differences, and the influencing factors of postoperative recurrence in elderly patients with kidney stones were obtained. A receiver operating characteristic (ROC) curve was drawn to analyse the value of the factors in predicting postoperative recurrence in patients with kidney stones.

A total of 123 elderly patients with renal calculi were enrolled. The patients were divided according to the presence or absence of stone recurrence into the recurrence group (25 cases, 20.33%) and the non-recurrence group (98 cases, 79.67%). Postoperative water intake, excessive intake of animal protein, exercise and postoperative complications significantly differed between the recurrence group and the non-recurrence group (p < 0.001). Logistic regression analysis showed that the above-mentioned indicators were the influencing factors of postoperative recurrence. The area under the curve (AUC) values of postoperative water intake (AUC = 0.767), animal protein intake (AUC = 0.752), exercise (AUC = 0.707) and postoperative complications (AUC = 0.727) were statistically significant, and they were identified as the most important factors with high sensitivity and specificity and were of high value in predicting postoperative recurrence of renal calculi.

Elderly patients with kidney stones are prone to recurrence after surgery. Influencing factors should be given attention, and corresponding measures should be formulated for intervention as soon as possible.

Archivos espanoles de urologia. 2024 Aug [Epub]

Lingyan Ding, Siyang Xu, Xinfeng Chen, Cheng Shen, Hua Zhu, Bing Zheng, Wei Zhang, Chunmei Shi

Urinary Surgery, Affiliated Hospital 2 of Nantong University, 226006 Nantong, Jiangsu, China.

PubMed http://www.ncbi.nlm.nih.gov/pubmed/39238302

Login to update email address, newsletter preferences and use bookmarks.

• Forgot Password?

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

The PMC website is updating on October 15, 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• v.2018; 2018

Kidney Stone Disease: An Update on Current Concepts

Tilahun alelign.

1 Department of Microbial, Cellular and Molecular Biology, College of Natural Sciences, Addis Ababa University, P.O. Box 1176, Addis Ababa, Ethiopia

2 Department of Biology, Debre Birhan University, P.O. Box 445, Debre Birhan, Ethiopia

Beyene Petros

Kidney stone disease is a crystal concretion formed usually within the kidneys. It is an increasing urological disorder of human health, affecting about 12% of the world population. It has been associated with an increased risk of end-stage renal failure. The etiology of kidney stone is multifactorial. The most common type of kidney stone is calcium oxalate formed at Randall's plaque on the renal papillary surfaces. The mechanism of stone formation is a complex process which results from several physicochemical events including supersaturation, nucleation, growth, aggregation, and retention of urinary stone constituents within tubular cells. These steps are modulated by an imbalance between factors that promote or inhibit urinary crystallization. It is also noted that cellular injury promotes retention of particles on renal papillary surfaces. The exposure of renal epithelial cells to oxalate causes a signaling cascade which leads to apoptosis by p38 mitogen-activated protein kinase pathways. Currently, there is no satisfactory drug to cure and/or prevent kidney stone recurrences. Thus, further understanding of the pathophysiology of kidney stone formation is a research area to manage urolithiasis using new drugs. Therefore, this review has intended to provide a compiled up-to-date information on kidney stone etiology, pathogenesis, and prevention approaches.

1. Introduction

1.1. overview of kidney stones.

Kidney stones are mainly lodged in the kidney(s) [ 1 ]. Mankind has been afflicted by urinary stones since centuries dating back to 4000 B.C. [ 2 ], and it is the most common disease of the urinary tract. The prevention of renal stone recurrence remains to be a serious problem in human health [ 3 ]. The prevention of stone recurrence requires better understanding of the mechanisms involved in stone formation [ 4 ]. Kidney stones have been associated with an increased risk of chronic kidney diseases [ 5 ], end-stage renal failure [ 3 , 6 ], cardiovascular diseases [ 7 , 8 ], diabetes, and hypertension [ 9 ]. It has been suggested that kidney stone may be a systemic disorder linked to the metabolic syndrome. Nephrolithiasis is responsible for 2 to 3% of end-stage renal cases if it is associated with nephrocalcinosis [ 10 ].

The symptoms of kidney stone are related to their location whether it is in the kidney, ureter, or urinary bladder [ 11 ]. Initially, stone formation does not cause any symptom. Later, signs and symptoms of the stone disease consist of renal colic (intense cramping pain), flank pain (pain in the back side), hematuria (bloody urine), obstructive uropathy (urinary tract disease), urinary tract infections, blockage of urine flow, and hydronephrosis (dilation of the kidney). These conditions may result in nausea and vomiting with associated suffering from the stone event [ 12 ]. Thus, the treatment and time lost from work involves substantial cost imposing an impact on the quality of life and nation's economy.

1.2. Epidemiology of Kidney Stones

Globally, kidney stone disease prevalence and recurrence rates are increasing [ 13 ], with limited options of effective drugs. Urolithiasis affects about 12% of the world population at some stage in their lifetime [ 14 ]. It affects all ages, sexes, and races [ 15 , 16 ] but occurs more frequently in men than in women within the age of 20–49 years [ 17 ]. If patients do not apply metaphylaxis, the relapsing rate of secondary stone formations is estimated to be 10–23% per year, 50% in 5–10 years, and 75% in 20 years of the patient [ 15 ]. However, lifetime recurrence rate is higher in males, although the incidence of nephrolithiasis is growing among females [ 18 ]. Therefore, prophylactic management is of great importance to manage urolithiasis.

Recent studies have reported that the prevalence of urolithiasis has been increasing in the past decades in both developed and developing countries. This growing trend is believed to be associated with changes in lifestyle modifications such as lack of physical activity and dietary habits [ 19 – 21 ] and global warming [ 16 ]. In the United States, kidney stone affects 1 in 11 people [ 22 ], and it is estimated that 600,000 Americans suffer from urinary stones every year. In Indian population, about 12% of them are expected to have urinary stones and out of which 50% may end up with loss of kidney functions [ 23 ].

2. The Urinary System and Stones

The urinary filtrate is formed in the glomerulus and passes into the tubules where the volume and content are altered by reabsorption or secretions. Most solute reabsorption occurs in the proximal tubules, whereas fine adjustments to urine composition take place in the distal tubule and collecting ducts. The loop of Henle serves to concentrate urine composed of 95% water, 2.5% urea, 2.5% mixture of minerals, salts, hormones, and enzymes. In the proximal tubules, glucose, sodium, chloride, and water are reabsorbed and returned to the blood stream along with essential nutrients such as amino acids, proteins, bicarbonate, calcium, phosphate, and potassium. In the distal tubule, the salt and acid-base balance of blood is regulated [ 24 ]. The location of stones may vary as indicated in Figure 1 .

Kidney stone locations in the urinary system. (a) Adopted from [ 25 ]. (b) Adopted from [ 26 ].

3. Types of Kidney Stones

The chemical composition of kidney stones depends on the abnormalities in urine composition of various chemicals. Stones differ in size, shape, and chemical compositions (mineralogy) [ 27 ]. Based on variations in mineral composition and pathogenesis, kidney stones are commonly classified into five types as follows [ 28 ].

3.1. Calcium Stones: Calcium Oxalate and Calcium Phosphate

Calcium stones are predominant renal stones comprising about 80% of all urinary calculi [ 29 ]. The proportion of calcium stones may account for pure calcium oxalate (CaOx) (50%), calcium phosphate (CaP, termed as apatite) (5%), and a mixture of both (45%) [ 30 ]. The main constituent of calcium stones is brushite (calcium hydrogen phosphate) or hydroxyapatite [ 31 , 32 ]. Calcium oxalate is found in the majority of kidney stones and exists in the form of CaOx monohydrate (COM, termed as mineral names: whewellite, CaC 2 O 4 ·H 2 O), and CaOx dihydrate (COD, weddellite, CaC 2 O 4 ·2H 2 O), or as a combination of both which accounts for greater than 60% [ 33 ]. COM is the most thermodynamically stable form of stone. COM is more frequently observed than COD in clinical stones [ 34 ].

Many factors contribute to CaOx stone formation such as hypercalciuria (resorptive, renal leak, absorptive, and metabolic diseases), hyperuricosuria, hyperoxaluria, hypocitraturia, hypomagnesuria, and hypercystinuria [ 35 ]. Mostly, urinary pH of 5.0 to 6.5 promotes CaOx stones [ 36 ], whereas calcium phosphate stones occur when pH is greater than 7.5 [ 11 ]. The recurrence of calcium stone is greater than other types of kidney stones.

3.2. Struvite or Magnesium Ammonium Phosphate Stones

Struvite stones occur to the extent of 10–15% and have also been referred to as infection stones and triple phosphate stones. It occurs among patients with chronic urinary tract infections that produce urease, the most common being Proteus mirabilis and less common pathogens include Klebsiella pneumonia , Pseudomonas aeruginosa, and Enterobacter [ 1 , 28 , 29 ]. Urease is necessary to split/cleave urea to ammonia and CO 2, making urine more alkaline which elevates pH (typically > 7). Phosphate is less soluble at alkaline versus acidic pH, so phosphate precipitates on to the insoluble ammonium products, yielding to a large staghorn stone formation [ 37 ]. Women's are likely to develop this type of stone than the male. Escherichia coli is not capable of splitting urea and is not associated with struvite stones [ 38 ].

3.3. Uric Acid Stones or Urate

This accounts approximately for 3–10% of all stone types [ 1 , 29 ]. Diets high in purines especially those containing animal protein diet such as meat and fish, results in hyperuricosuria, low urine volume, and low urinary pH (pH < 5.05) exacerbates uric acid stone formation [ 11 , 28 , 39 ]. Peoples with gouty arthritis may form stones in the kidney(s). The most prevalent cause of uric acid nephrolithiasis is idiopathic [ 38 ], and uric acid stones are more common in men than in women.

3.4. Cystine Stones

These stones comprise less than 2% of all stone types. It is a genetic disorder of the transport of an amino acid and cystine. It results in an excess of cystinuria in urinary excretions [ 1 , 29 ], which is an autosomal recessive disorder caused by a defect in the rBAT gene on chromosome 2 [ 40 ], resulting in impaired renal tubular absorption of cystine or leaking cystine into urine. It does not dissolve in urine and leads to cystine stone formation [ 11 ]. People who are homozygous for cystinuria excrete more than 600 millimole insoluble cystine per day [ 28 ]. The development of urinary cystine is the only clinical manifestation of this cystine stone disease [ 40 ].

3.5. Drug-Induced Stones

This accounts for about 1% of all stone types [ 1 ]. Drugs such as guaifenesin, triamterene, atazanavir, and sulfa drugs induce these stones. For instance, people who take the protease inhibitor indinavir sulphate, a drug used to treat HIV infection, are at risk of developing kidney stones [ 28 ]. Such lithogenic drugs or its metabolites may deposit to form a nidus or on renal calculi already present. On the other hand, these drugs may induce the formation of calculi through its metabolic action by interfering with calcium oxalate or purine metabolisms [ 38 ].

4. Kidney Stone Compositions

The chemical compositions of urinary stones include crystals and noncrystalline phases or the organic material (the matrix). The organic matrix of urinary stones consists of macromolecules such as glycosaminoglycans (GAG's), lipids, carbohydrates, and proteins. These molecules play a significant role by promoting or inhibiting the processes of kidney stone development ( Table 1 ). The main components of the stone matrix are proteins (64%), nonamino sugars (9.6%), hexosamine as glucosamine (5%), water (10%), and inorganic ash (10.4%). The matrix acts as a template participating in the assembly of kidney stones. The matrix of all stones contains phospholipids (8.6%) of the total lipid, which in turn represents about 10.3% of stone matrix. Cell membrane phospholipids, as part of organic matrix, promote the formation of calcium oxalate and calcium phosphate stones [ 41 ]. Albumin is the major component of the matrix of all stone types [ 42 ].

Urinary stone matrix protein modulators of crystallization in nephrolithiasis [ 34 , 41 ].

I: inhibitor; P: promoter; “—”: no effect.

Brushite stone is a hard phosphate mineral with an increasing incidence rate, and a quarter of calcium phosphate (CaP) patients form stones containing brushite [ 43 ]. In the urinary tract, CaP may be present in the form of hydroxyapatite, carbonate apatite, or brushite (calcium monohydrogen phosphate dihydrate, CaHPO4·2H2O). Brushite is resistant to shock wave and ultrasonic lithotripsy treatment [ 44 ].

4.1. Etiology of Kidney Stones

Formation of kidney stones (calculogenesis) is a complex and multifactorial process including intrinsic (such as age, sex, and heredity) and extrinsic factors such as geography, climate, dietary, mineral composition, and water intake [ 15 ]. A summary of possible causes of kidney stone formation is shown in Table 2 .

Risk factors associated with kidney stone formations.

5. Mechanisms of Renal Stone Formation

The pathogenesis of kidney stone or biomineralization is a complex biochemical process which remains incompletely understood [ 41 ]. Renal stone formation is a biological process that involves physicochemical changes and supersaturation of urine. Supersaturated solution refers to a solution that contains more of dissolved material than could be dissolved by the solvent under normal circumstances [ 34 ]. As a result of supersaturation, solutes precipitate in urine leads to nucleation and then crystal concretions are formed. That is, crystallization occurs when the concentration of two ions exceeds their saturation point in the solution [ 55 ]. The transformation of a liquid to a solid phase is influenced by pH and specific concentrations of excess substances. The level of urinary saturation with respect to the stone-forming constituents like calcium, phosphorus, uric acid, oxalate, cystine, and low urine volume are risk factors for crystallization [ 1 , 56 ]. Thus, crystallization process depends on the thermodynamics (that leads to nucleation) and kinetics (which comprises the rates of nucleation or crystal growth) of a supersaturated solution [ 57 ]. Therefore, lithiasis can be prevented by avoiding supersaturation.

However, it should be noted that stone formation is usually dependent on the level of imbalance between urinary inhibitors and promoters of crystallization. All stones share similar events with respect to the mineral phase of stone formation. But, the sequence of events leading to stone formation differs depending on the type of stone and urine chemistry. For instance, crystallization of calcium-based stones (calcium oxalate or calcium phosphate) occurs in supersaturated urine if it is with low concentrations of inhibitors. Uric acid interferes the solubility of calcium oxalate and promotes CaOx stone formation. In healthy controls, crystallization process is opposed by inhibitory substances and gets safe [ 1 ]. The sequence of events that trigger stone formation includes nucleation, growth, aggregation, and retention of crystals within the kidneys [ 27 , 58 ].

5.1. Crystal Nucleation

The first step in the formation of kidney stone begins by the formation of nucleus (termed as nidus) from supersaturated urine retained inside the kidneys [ 11 , 42 ]. In a supersaturated liquid, free atoms, ions, or molecules start forming microscopic clusters that precipitate when the bulk free energy of the cluster is less than that of the liquid. For example, charged soluble molecules such as calcium and oxalate combine to form calcium oxalate crystals and become insoluble [ 34 ]. Nucleation may be formed in the kidney through free particle or fixed particle mechanism [ 26 , 34 ]. In supersaturated solutions, if promoters exceed that of inhibitors, nucleation starts [ 34 ].

Once a nucleus is created (and/or if it is anchored), crystallization can occur at lower chemical pressure than required for the formation of the initial nucleus. Existing epithelial cells, urinary casts, RBCs, and other crystals in urine can act as nucleating centers in the process of nuclei formation termed as heterogeneous nucleation [ 41 ]. The organic matrix, mucopolysaccharide acts as a binding agent by increasing heterogeneous nucleation and crystal aggregation [ 59 ]. On the other hand, nanobacteria is claimed to form apatite structures serving as a crystallization center for stone formation [ 60 ]. The whole process potentiates stone formation. The role of oxalate-degrading bacteria, such as Oxalobacter formigenes , in CaOx stone formation is a subject of current research [ 61 ]. Thus, treatment which targets the process of nucleation intervention is one of the best approaches to control kidney stone.

5.2. Crystal Growth

Crystals in urine stick together to form a small hard mass of stone referred as crystal growth. Stone growth is accomplished through aggregation of preformed crystals or secondary nucleation of crystal on the matrix-coated surface [ 62 ]. Once a nidus has achieved, the overall free energy is decreased by adding new crystal components to its surface. The total free energy of the cluster is increased by the surface energy. The process of stone growth is slow and requires longer time to obstruct the renal tubules [ 34 ]. From organic matrix, mainly Tamm–Horsfall protein and osteopontin are promoters of CaOx stone formation [ 13 ]. Under in vitro study, crystals induced in human urine demonstrated an intimate association between calcium-containing crystals and organic matrix (lipids and proteins). Lipids of cellular membranes are basically believed to involve in nucleation of crystals [ 63 ].

5.3. Crystal Aggregation

The process whereby a small hard mass of a crystal in solution sticks together to form a larger stone is called aggregation. All models of CaOx urolithiasis concede that crystal aggregation is probably involved in crystal retention within the kidneys [ 41 ]. Crystal aggregation is considered to be the most critical step in stone formation.

5.4. Crystal-Cell Interaction

The attachment of grown crystals with the renal tubule lining of epithelial cells is termed as crystal retention or crystal-cell interaction [ 41 , 64 ]. In individuals with hyperoxaluria, renal tubular epithelial cells were injured due to exposure to high oxalate concentrations or sharp calcium oxalate monohydrate (COM) crystals [ 10 , 65 , 66 ]. Crystal-cell interaction results in the movement of crystals from basolateral side of cells to the basement membrane [ 10 ]. Then, crystals could be taken into cells and anchored to the basement membrane of the kidneys [ 66 ]. The interaction of COM crystals with the surface of renal epithelial cells could be a critical initiating event in nephrolithiasis. An increased retention force between the crystal and injured renal tubule epithelium cells promotes CaOx crystallization [ 67 ]. Most of the crystals attached to epithelial cells are thought to be digested by macrophages and/or lysosomes inside cells and then discharged with urine [ 66 ].

Following renal tubular cell injury, cellular degradation produces numerous membrane vesicles which are nucleators of calcium crystals as supported by in vitro and in vivo studies [ 41 ]. Injured cells release substances like renal prothrombin fragment-1 or other anionic proteins which induce COM crystal agglomeration [ 68 ]. Reactive oxygen species is thought to be one of the factors involved in renal cell injury [ 69 ]. Thus, reduction of renal oxidative stress could be an effective treatment option.

Injured cells potentiate to invert its cell membrane which is anionic to the urinary environment and acts as site of crystal adherence. COM crystals have stronger affinity of attachment towards the inverted anionic membrane [ 69 ], than calcium oxalate dihydrate (COD) crystals [ 70 ]. On the other hand, deposition of COM crystal was observed in Madin–Darby canine kidney epithelial cells (MDCK cells), than at proximal tubular epithelial cells derived from pig kidney (LLC-PK1 cells) study models [ 71 ]. This preferential difference may be due to the presence of a binding molecule such as hyaluronan on Madin–Darby canine kidney epithelial cells for COM crystal attachment [ 67 ]. Although the detailed mechanisms of crystal-cell interaction remain unexplored, one of the best ways to treat urolithiasis is to control crystal-cell retentions.

5.5. Endocytosis of CaOx Crystals

Endocytosis or engulfment of crystals by renal tubular cells is the earliest process in the formation of kidney stones. Studies on tissue culture crystal-cell interactions indicated that COM crystals rapidly adhere to microvilli on the cell surface and subsequently internalized. Polyanion molecules present in tubular fluid/urine such as glycosaminoglycans, glycoproteins, and citrate may coat crystals and inhibit the binding of COM crystals to cell membrane [ 41 ]. For example, Tamm–Horsfall glycoproteins (THP) have a dual biological role in stone formation. Lieske et al. [ 72 ] reported that THP may promote renal stone formation by initiating the interaction of COM crystals with distal tubular cells of the nephron. Another study revealed that, upon lowering pH and raising ionic strength, THP's viscosity increases which exhibits high tendency of polymerization and fails to inhibit crystallization. Moreover, THP becomes a strong promoter of crystallization in the presence of additional calcium ions [ 73 ]. In contrast, THP is thought to protect against COM stone formation by inhibiting COM aggregation when it is at high pH and low ionic strength as reported by Hess [ 73 ]. COM aggregation assays revealed that desialylated THP promoted COM aggregation, while normal THP inhibited aggregation [ 74 ]. Similar reports revealed that THP may inhibit calcium oxalate crystal aggregation, whereas uromodulin may promote aggregation [ 75 ]. Inactivating the THP gene in mouse embryonic stem cells results in spontaneous formation of calcium crystals in adult kidneys. This is a convincing evidence that THP is a critical urinary inhibitor of human nephrolithiasis [ 76 ].

Various cellular and extracellular events are involved during stone formation. Modulators targeting the steps from supersaturation to crystal retention may be a potential means to block stone formation. Similarly, the blockage of crystal binding molecules (such as osteopontin, hyaluronic acid, sialic acid, and monocyte chemoattractant protein-1) expressed on epithelial cell membranes may be an alternative approach to prevent stone formation [ 41 ]. Experimental findings demonstrated that stone calcification is triggered by reactive oxygen species (ROS) and the development of oxidative stress [ 77 ]. In vitro [ 78 , 79 ] and in vivo [ 80 , 81 ] studies have demonstrated that CaOx crystals are toxic for renal epithelial cells that produce injury and renal cell death. Similarly, an exposure to hypercalciuria produces cellular injury and ROS-induced lipid peroxidation which stimulates calcium oxalate deposition [ 82 ]. The pathophysiology of urinary stone formation is incompletely understood. A summary of the various steps involved in stone formation is shown below ( Figure 2 ).

Schematic representation of the various events of kidney stone formation.

5.6. Cell Injury and Apoptosis

Exposure to high levels of oxalate or CaOx crystals induces epithelial cellular injury, which is a predisposing factor to subsequent stone formation [ 83 , 84 ]. CaOx crystal depositions in the kidneys upregulate the expression and synthesis of macromolecules that can promote inflammation [ 85 ]. Crystals may be endocytosed by cells or transported to the interstitium. It has been suggested that injured cells develop a nidus which promotes the retention of particles on the renal papillary surface [ 86 ]. In individuals with severe primary hyperoxaluria, renal tubular cells are injured and crystals become attached to them [ 66 ]. The addition of CaOx crystals onto Madin–Darby canine kidney (MDCK) cell lines showed an increase in the release of lysosomal enzymes, prostaglandin E2, and cytosolic enzymes [ 87 ]. A study on animal models also revealed that the administration of high concentrations of CaOx crystals or oxalate ions appears to be toxic causing renal tubular cell damage [ 41 ]. It has been suggested that oxalate increases the availability of free radicals by inhibiting enzymes responsible for their degradation. For instance, reactive oxygen species can damage the mitochondrial membrane and reduce its transmembrane potential. These events are known features of early process in apoptotic pathways [ 88 ].

The activation of p38 mitogen-activated protein kinase (p38 MAPK) signaling pathway regulates the expression of cellular proteins. The various extracellular stimuli or stresses like ultraviolet radiation and proinflammatory cytokines may activate p38 MAPK which results in phosphorylation and activation of transcription factors [ 89 ]. The exposure of renal cells to oxalate increases an altered gene expression that induces apoptosis signaling cascades [ 88 ]. A study revealed that the exposure of HK-2 cells to increased oxalate levels results in an increased transcriptional activation of IL-2R beta mRNA and consequently increases IL-2R beta protein levels which drive cellular changes like induction of inflammation. Oxalate-induced activation may trigger p38 MAPK signaling by acting on cell membranes, although the exact mechanisms have not been established [ 90 ].

Apoptosis at the level of renal tubular cells may lead to stone formation through cellular demise and postapoptotic necrosis which could promote calcium crystal aggregation and growth. This fact has been supported by in vitro study on MDCK cells being exposed to oxalate ions [ 91 ]. However, it has to be noted that some cells did not respond to oxalate injury. This may be due to the fact that changes in gene expression could protect from apoptosis and then inhibit from lithiasis [ 35 ]. These findings highlight the need for future studies clarifying novel biochemical targets of kidney stone formation and the utility of p38 MAPK inhibitors in preventing stone formation.

5.7. Genetic Basis of Kidney Stone Formation

Environmental factors interacting with underlying genetic factors cause rare stone disease [ 92 ]. The production of promoters and inhibitors of crystallization depends on proper functioning of the renal epithelial cells. Cellular dysfunction affects the supersaturation of urinary excretion by influencing ions such as calcium, oxalate, and citrate [ 93 ]. Some genetic defects which lead to stone formation are shown in Table 3 .

Gene involved in hypercalciuria, gene products, and renal phenotype [ 93 ].

5.8. Randall's Plaques

Randall's plaques appear to be the precursor's origin of urinary stone development although it is unclear whether it involves in all stone types or not [ 62 ]. Moreover, the pathogenesis of Randall's plaque itself is not clearly known [ 94 ]. The majority of CaOx stones are found to be attached with renal papillae at the sites of Randall's plaque [ 26 ]. It is located at the interstitial basement membrane in loop of Henle [ 95 , 96 ]. Calcium phosphate (apatite), and purine crystal compositions were identified in plaques, whereas apatite is dominant [ 97 ]. Initially, calcium phosphate crystals and organic matrix are deposited along the basement membranes of the thin loops of Henle and extend further into the interstitial space to the urothelium, constituting the so-called Randall plaques. Evidence suggests that a primary interstitial apatite crystal formation secondarily leads to CaOx stone formation [ 13 ]. In supersaturated urine, crystals adhere to the urothelium which may enhance subsequent stone growth [ 98 ].

Due to renal cell injury, plaque is exposed to supersaturated urine. Renal epithelial cell damage (degradation) products promote heterogeneous nucleation and promotes crystal adherence in renal cells. Randall plaque calcification is triggered by oxidative stress. Cells may express molecules at distal and collecting tubules which act as crystal binding sites such as phosphatidylserine, CD44, osteopontin, and hyaluronan [ 27 , 99 ]. Renal epithelial cells of the loop of Henle or collecting ducts produce membrane vesicles at the basal side which leads to plague formation [ 77 ]. Thus, apatite crystal deposits have been proposed to act as nidus for CaOx stone formation by attachment on further matrix molecules [ 13 , 77 ]. However, the driving forces in plaque formation and the involved matrix molecules remain elusive.

Kidney stones are either attached to the renal papillae or found freely [ 100 ]. According to the fixed particle pathway, the beginning of calcium phosphate (CaP) deposition in the interstitium establishes a nucleus for CaOx formation. CaP formed in the basement membrane of the loops of Henle, the inner medullary collecting ducts, and ducts of Bellini serves as an attachment site for stone development. Idiopathic stone formers develop CaOx attached to fixed sites of interstitial plaque [ 26 ]. Stones of the distal tubular acidosis attach to plugs protruding from dilated ducts of Bellini, whereas cystinuria stones do not attach to the renal plagues (found freely) [ 26 ]. CaP, uric acid, or cystine crystals formed in the renal tubules plug at the terminal collecting ducts. When mineralization reaches the renal papillary surface, plaques rupture exposing CaP crystals to the pelvic urine. Then, urinary macromolecules deposit over the exposed CaP crystals and promote CaOx deposition on CaP [ 4 ].

5.9. Kidney Stone Inhibitors and Promoters

Inhibitors are substances which decrease the initiation of supersaturation, nucleation, crystal growth, rate of aggregation, or any other processes required to stone formation [ 34 ]. Normally, urine contains chemicals that prevent crystal formation. Inhibitors in urine includes small organic anions such as citrate, small inorganic anions such as pyrophosphates, multivalent metallic cations such as magnesium, or macromolecules such as osteopontin, glycosaminoglycans, glycoproteins, urinary prothrombin fragment-1, and Tamm–Horsfall proteins [ 41 , 62 ]. These inhibitors do not seem to work equally for everyone; therefore, some people form stones. But, if crystals formed remain tiny, usually it travels through the urinary tract and passes out from the body with urine splash without being noticed. Inhibitors may act either directly by interacting with crystal or indirectly by influencing the urinary environment [ 42 ]. When inhibitory compounds adsorb onto the surface of the crystal, it inhibits nucleation, crystal growth, aggregation, or crystal-cell adherence.

In contrast, promoters are substances which facilitate stone formation by various mechanisms [ 62 ]. Some of the promoters include cell membrane lipids (phospholipids, cholesterol, and glycolipids) [ 42 ], calcitriol hormone enhancement via parathyroid hormone stimulation [ 101 ], oxalate, calcium, sodium, cystine, and low urine volume [ 34 ]. Among recurrent stone formers, urinary oxalate excretion was found to be higher, whereas citrate excretion was lower [ 102 ]. Studies indicated that oxalate can increase chloride, sodium, and water reabsorption in the proximal tubule and activate multiple signaling pathways in renal epithelial cells [ 103 ]. In general, an imbalance between urinary stone inhibitors and promoters has been suggested to be the cause for stone formation [ 34 ]. A list of substances generally considered to inhibit or promote stone formation process is shown in Table 1 .

6. Preventive Options for Urolithiasis

Effective kidney stone prevention depends upon addressing the cause of stone formation. Generally, to prevent the first episodes of kidney stone formation or its secondary episodes, proper management of diet and the use of medications is required. Primary prevention of kidney stone disease via dietary intervention is low-cost public health initiative with massive societal implications. Thus, nutritional management is the best preventive strategy against urolithiasis [ 104 ].

Regardless of the underlying etiology and drug treatment of the stone disease, patients should be instructed to increase their water intake in order to maintain a urine output of at least 2 liter per day [ 49 ]. A simple and most important lifestyle change to prevent stone disease is to drink more water/liquids. Enough fluid intake reduces urinary saturation and dilutes promoters of CaOx crystallization. Dietary recommendations should be adjusted based on individual metabolic abnormalities. For absorptive hyperoxaluria, low oxalate diet and increased dietary calcium intake are recommended [ 61 ].

A high sodium intake boosts stone risk by reducing renal tubular calcium reabsorption and increasing urinary calcium [ 105 ]. Restriction of animal proteins is also encouraged since animal proteins provide an increased acid load because of its high content of sulfur-containing amino acids. Thus, high protein intake reduces urine pH and the level of citrate and enhances urinary calcium excretion via bone reabsorption. Therefore, if you have very acidic urine, you may need to eat less meat, fish, and poultry and avoid food with vitamin D [ 106 ]. Instead, an increase intake of fruits and vegetables rich in potassium is recommended [ 49 ].

People who form calcium stones used to be told to avoid dairy products and other foods with high calcium content. However, persons with a tendency of kidney stone formation should not be advised to restrict calcium intake unless it has been known that he/she has an excessive use of calcium [ 107 ]. A reduced intake of calcium leads to an increased intestinal absorption of oxalate, which itself may account for an increased risk of stone formation. Calcium supplements may reduce oxalate absorption because the calcium binds dietary oxalate in the gut lumen. However, the benefit of taking calcium pills is controversial. Vitamin C has been implicated in stone formation because of in vivo conversion of ascorbic acid to oxalate. Therefore, a limitation of vitamin C supplementation is recommended [ 105 ].

For prevention of calcium oxalate, cystine, and uric acid stones, urine should be alkalinized by eating a diet high in fruits and vegetables, taking supplemental or prescription citrate, or drinking alkaline mineral waters. For uric acid stone formers, gout needs to be controlled, and for cystine stone formers, sodium and protein intakes need to restricted. For prevention of calcium phosphate and struvite stones, urine should be acidified. For struvite stones, acidifying the urine is the single most important step [ 108 ]. Patients must receive careful follow-up to be sure that the infection has cleared. However, the current treatment modalities are not efficient to prevent urolithiasis, and further research is required.

7. Conclusion

Despite considerable improvements in the development of new therapies for the management of urinary stones, the incidence of urolithiasis is increasing worldwide. Many aspects of renal stone formation remain unclear. However, it is certain that renal cell injury, crystal retention, cell apoptosis, Randall's plaque, and associated stone inhibitors or promoters play important roles for kidney stone formation. These seem to be critical targets that lead to developing a novel strategy to prevent kidney stone disease and drugs against kidney stones. In addition, the identification of novel treatment targets on the basis of molecular and cellular alterations in relation to stone formation will help develop better drugs. Moreover, better understanding of the mechanisms of urolithiasis associated with stone inhibitors or promoters will be critical for stone-removing medications. Furthermore, understanding the underlying pathophysiology, pathogenesis, and genetic basis of kidney stone formation will hopefully lead to discover novel drugs and strategies to manage urolithiasis in the near future.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

• Peer Exchange
• Peers and Perspectives
• Case-Based Peer Perspectives
• Urology Times Journal

• Conferences
• Publications

Challenging cases in urology: A case of hydronephrosis, sepsis, and pain

A 61-year-old female complains of acute-onset right-sided flank and abdominal pain. She describes associated nausea, but denied urinary symptoms, hematuria, fever, or chills.

Sara Valente, MD, is a fourth-year urology resident at the University of Connecticut School of Medicine, Farmington.

“Challenging Cases in Urology” is a new Urology Times section in which residents from the nation’s leading urology programs present their toughest cases and how they ultimately managed them. Cases inform readers of the problem-solving process and provide a lesson from the authors’ experience.

A 61-year-old female presented to the emergency department with a complaint of acute-onset right-sided flank and abdominal pain that awoke her from sleep. She described associated nausea, but denied urinary symptoms, hematuria, fever, or chills. One month ago she had been treated for a presumed urinary tract infection after presenting with dysuria and urinary frequency. She also reported intermittent abdominal pain dating back approximately 2 months.

Her past medical and surgical history was notable for a right-sided UPJ repair in 1978 as well as a history of nephrolithiasis and right-sided ESWL. She did not know the stone composition. Additional surgical history included only a previous tonsillectomy.

Examination

Evaluation in the emergency department revealed a leukocytosis of 16.7 thou/µL, a normal serum creatinine, and a clean catch urinalysis with 1 WBC/HPF, 4 RBC/HPF, nitrite positive, and leukocyte esterase negative. Complete metabolic panel and liver function tests were all within normal limits. A CT scan demonstrated moderate right-sided hydronephrosis, symmetric nephrograms, and a small calcification that appeared to be within the distal right ureter (figures 1 and 2).

FDA Advisory Committee meets to discuss NDA of oral sulopenem for uUTI

Follow Urology Times' coverage of the full-day FDA advisory committee meeting regarding Iterum Therapeutics' new drug application for oral sulopenem for uncomplicated urinary tract infection.

Dr. Bristol Whiles discusses her experience as a urology resident

On this podcast, Bristol Whiles, MD, discusses her urology residency experience at the Kansas University School of Medicine Department of Urology.

Ultrasonic propulsion for clearance of residual kidney stone fragments reduces relapse

In total, 8 of 40 patients who underwent ultrasonic propulsion experienced relapse, compared with 21 of 42 patients who underwent observation.

Destigmatizing Urology: Dr. Winter discusses STDs/STIs

“We need to just approach STIs as the medical conditions they are because until we do, and until we take those stigmas away, people will truly have resistance to discussing it, hesitance to discussing it, and resistance to diagnosis and intervention,” says Ashley Winter, MD.

Trial launches of gene editing therapy for primary hyperoxaluria type 1

In total, the trial plans to enroll 7 patients with PH1 through a single center in China.

Data suggest Break Wave lithotripsy is safe, effective for urolithiasis

Any stone fragmentation was achieved in 88% of patients.

2 Commerce Drive Cranbury, NJ 08512

609-716-7777

Information

• Author Services

Initiatives

You are accessing a machine-readable page. In order to be human-readable, please install an RSS reader.

All articles published by MDPI are made immediately available worldwide under an open access license. No special permission is required to reuse all or part of the article published by MDPI, including figures and tables. For articles published under an open access Creative Common CC BY license, any part of the article may be reused without permission provided that the original article is clearly cited. For more information, please refer to https://www.mdpi.com/openaccess .

Feature papers represent the most advanced research with significant potential for high impact in the field. A Feature Paper should be a substantial original Article that involves several techniques or approaches, provides an outlook for future research directions and describes possible research applications.

Feature papers are submitted upon individual invitation or recommendation by the scientific editors and must receive positive feedback from the reviewers.

Editor’s Choice articles are based on recommendations by the scientific editors of MDPI journals from around the world. Editors select a small number of articles recently published in the journal that they believe will be particularly interesting to readers, or important in the respective research area. The aim is to provide a snapshot of some of the most exciting work published in the various research areas of the journal.

Original Submission Date Received: .

• Active Journals
• Find a Journal
• Journal Proposal
• Proceedings Series
• For Authors
• For Reviewers
• For Editors
• For Librarians
• For Publishers
• For Societies
• For Conference Organizers
• Open Access Policy
• Institutional Open Access Program
• Special Issues Guidelines
• Editorial Process
• Research and Publication Ethics
• Article Processing Charges
• Testimonials
• Preprints.org
• SciProfiles
• Encyclopedia

Article Menu

• Subscribe SciFeed
• Google Scholar
• on Google Scholar
• Table of Contents

Find support for a specific problem in the support section of our website.

Please let us know what you think of our products and services.

Visit our dedicated information section to learn more about MDPI.

JSmol Viewer

Successful treatment of multiple large intrarenal stones in a 2-year-old boy using a single-use flexible ureteroscope and high-power laser settings.

1. Introduction

2. case description, 2.1. first referral and further investigation, 2.2. surgical technique and follow-up, 3. discussion, 4. conclusions, author contributions, institutional review board statement, informed consent statement, data availability statement, conflicts of interest, abbreviation list.

• Tasian, G.E.; Copelovitch, L. Evaluation and medical management of kidney stones in children. J. Urol. 2014 , 192 , 1329–1336. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Erkurt, B.; Caskurlu, T.; Atis, G.; Gurbuz, C.; Arikan, O.; Pelit, E.S.; Altay, B.; Erdogan, F.; Yildirim, A. Treatment of renal stones with flexible ureteroscopy in preschool age children. Urolithiasis 2014 , 42 , 241–245. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• European Association Urology. EAU Guidelines, Edn. In Proceedings of the EAU Annual Congress Milan 2023, Milan, Italy, 10–13 March 2023. [ Google Scholar ]
• Ishii, H.; Griffin, S.; Somani, B.K. Flexible ureteroscopy and lasertripsy (FURSL) for paediatric renal calculi: Results from a systematic review. J. Pediatr. Urol. 2014 , 10 , 1020–1025. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Tonyali, S.; Haberal, H.B.; Esperto, F.; Hamid, Z.; Tzelves, L.; Pietropaolo, A.; Emiliani, E. The Prime Time for Flexible Ureteroscopy for Large Renal Stones Is Coming: Is Percutaneous Nephrolithotomy No Longer Needed? Urol. Res. Pract. 2023 , 49 , 280–284. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Juliebo-Jones, P.; Ventimiglia, E.; Somani, B.K.; MS, A.E.; Gjengsto, P.; Beisland, C.; Ulvik, O. Single use flexible ureteroscopes: Current status and future directions. BJUI Compass 2023 , 4 , 613–621. [ Google Scholar ] [ CrossRef ]
• Carey, R.I.; Gomez, C.S.; Maurici, G.; Lynne, C.M.; Leveillee, R.J.; Bird, V.G. Frequency of ureteroscope damage seen at a tertiary care center. J. Urol. 2006 , 176 , 607–610, discussion 610. [ Google Scholar ] [ CrossRef ]
• Fang, L.; Xie, G.; Zheng, Z.; Liu, W.; Zhu, J.; Huang, T.; Lu, Y.; Cheng, Y. The Effect of Ratio of Endoscope-Sheath Diameter on Intrapelvic Pressure During Flexible Ureteroscopic Lasertripsy. J. Endourol. 2019 , 33 , 132–139. [ Google Scholar ] [ CrossRef ]
• Mille, E.; El-Khoury, E.; Haddad, M.; Pinol, J.; Charbonnier, M.; Gastaldi, P.; Dariel, A.; Merrot, T.; Faure, A. Comparison of single-use flexible ureteroscopes with a reusable ureteroscope for the management of paediatric urolithiasis. J. Pediatr. Urol. 2023 , 19 , 248.e1–248.e6. [ Google Scholar ] [ CrossRef ]
• Tiselius, H.G.; Andersson, A. Stone burden in an average Swedish population of stone formers requiring active stone removal: How can the stone size be estimated in the clinical routine? Eur. Urol. 2003 , 43 , 275–281. [ Google Scholar ] [ CrossRef ]
• Tasian, G.E.; Ross, M.E.; Song, L.; Sas, D.J.; Keren, R.; Denburg, M.R.; Chu, D.I.; Copelovitch, L.; Saigal, C.S.; Furth, S.L. Annual Incidence of Nephrolithiasis among Children and Adults in South Carolina from 1997 to 2012. Clin. J. Am. Soc. Nephrol. 2016 , 11 , 488–496. [ Google Scholar ] [ CrossRef ]
• Ingvarsdottir, S.E.; Indridason, O.S.; Palsson, R.; Edvardsson, V.O. Stone recurrence among childhood kidney stone formers: Results of a nationwide study in Iceland. Urolithiasis 2020 , 48 , 409–417. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Jabbar, F.; Asif, M.; Dutani, H.; Hussain, A.; Malik, A.; Kamal, M.A.; Rasool, M. Assessment of the role of general, biochemical and family history characteristics in kidney stone formation. Saudi J. Biol. Sci. 2015 , 22 , 65–68. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Baatiah, N.Y.; Alhazmi, R.B.; Albathi, F.A.; Albogami, E.G.; Mohammedkhalil, A.K.; Alsaywid, B.S. Urolithiasis: Prevalence, risk factors, and public awareness regarding dietary and lifestyle habits in Jeddah, Saudi Arabia in 2017. Urol. Ann. 2020 , 12 , 57–62. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Sulaiman, S.K.; Enakshee, J.; Traxer, O.; Somani, B.K. Which Type of Water Is Recommended for Patients with Stone Disease (Hard or Soft Water, Tap or Bottled Water): Evidence from a Systematic Review over the Last 3 Decades. Curr. Urol. Rep. 2020 , 21 , 6. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Rossi, M.; Barone, B.; Di Domenico, D.; Esposito, R.; Fabozzi, A.; D’Errico, G.; Prezioso, D. Correlation between Ion Composition of Oligomineral Water and Calcium Oxalate Crystal Formation. Crystals 2021 , 11 , 1507. [ Google Scholar ] [ CrossRef ]
• Zhang, Y.; Li, J.; Jiao, J.W.; Tian, Y. Comparative outcomes of flexible ureteroscopy and mini-percutaneous nephrolithotomy for pediatric kidney stones larger than 2 cm. Int. J. Urol. 2021 , 28 , 650–655. [ Google Scholar ] [ CrossRef ]
• Mahmoud, M.A.; Shawki, A.S.; Abdallah, H.M.; Mostafa, D.; Elawady, H.; Samir, M. Use of retrograde intrarenal surgery (RIRS) compared with mini-percutaneous nephrolithotomy (mini-PCNL) in pediatric kidney stones. World J. Urol. 2022 , 40 , 3083–3089. [ Google Scholar ] [ CrossRef ]
• Li, J.; Wang, W.; Du, Y.; Tian, Y. Combined use of flexible ureteroscopic lithotripsy with micro-percutaneous nephrolithotomy in pediatric multiple kidney stones. J. Pediatr. Urol. 2018 , 14 , 281.e1–281.e6. [ Google Scholar ] [ CrossRef ]
• Mosquera, L.; Pietropaolo, A.; Madarriaga, Y.Q.; de Knecht, E.L.; Jones, P.; Tur, A.B.; Griffin, S.; Somani, B.K. Is Flexible Ureteroscopy and Laser Lithotripsy the New Gold Standard for Pediatric Lower Pole Stones? Outcomes from Two Large European Tertiary Pediatric Endourology Centers. J. Endourol. 2021 , 35 , 1479–1482. [ Google Scholar ] [ CrossRef ]
• Candela, L.; Solano, C.; Castellani, D.; Teoh, J.Y.; Tanidir, Y.; Fong, K.Y.; Vaddi, C.; Mani Sinha, M.; Ragoori, D.; Somani, B.K.; et al. Comparing outcomes of thulium fiber laser versus high-power Holmium:YAG laser lithotripsy in pediatric patients managed with RIRS for kidney stones. A multicenter retrospective study. Minerva Pediatr. 2023 . [ Google Scholar ] [ CrossRef ]
• Garcia Rojo, E.; Traxer, O.; Vallejo Arzayus, D.M.; Castellani, D.; Ferreti, S.; Gatti, C.; Bujons, A.; Quiroz, Y.; Yuen-Chun Teoh, J.; Ragoori, D.; et al. Comparison of Low-Power vs High-Power Holmium Lasers in Pediatric Retrograde Intrarenal Surgery Outcomes. J. Endourol. 2023 , 37 , 509–515. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Quiroz Madarriaga, Y.; Badenes Gallardo, A.; Llorens de Knecht, E.; Motta Lang, G.; Palou Redorta, J.; Bujons Tur, A. Can cystinuria decrease the effectiveness of RIRS with high-power ho:yag laser in children? Outcomes from a tertiary endourology referral center. Urolithiasis 2022 , 50 , 229–234. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Bujons, A.; Millan, F.; Centeno, C.; Emiliani, E.; Sanchez Martin, F.; Angerri, O.; Caffaratti, J.; Villavicencio, H. Mini-percutaneous nephrolithotomy with high-power holmium YAG laser in pediatric patients with staghorn and complex calculi. J. Pediatr. Urol. 2016 , 12 , 253.e1–253.e5. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Berrettini, A.; Boeri, L.; Montanari, E.; Mogiatti, M.; Acquati, P.; De Lorenzis, E.; Gallioli, A.; De Marco, E.A.; Minoli, D.G.; Manzoni, G. Retrograde intrarenal surgery using ureteral access sheaths is a safe and effective treatment for renal stones in children weighing <20 kg. J. Pediatr. Urol. 2018 , 14 , 59.e1–59.e6. [ Google Scholar ] [ CrossRef ]
• Faure, A.; Paye Jaouen, A.; Demede, D.; Juricic, M.; Arnaud, A.; Garcia, C.; Charbonnier, M.; Abbo, O.; Botto, N.; Blanc, T.; et al. Safety and feasability of ureteroscopy for pediatric stone, in children under 5 Years (SFUPA 5): A French multicentric study. J. Pediatr. Urol. 2024 , 20 , 225.e1–225.e8. [ Google Scholar ] [ CrossRef ]

Click here to enlarge figure

Share and Cite

Tatanis, V.; Spinos, T.; Lamprinou, Z.; Kanna, E.; Mulita, F.; Peteinaris, A.; Achilleos, O.; Skondras, I.; Liatsikos, E.; Kallidonis, P. Successful Treatment of Multiple Large Intrarenal Stones in a 2-Year-Old Boy Using a Single-Use Flexible Ureteroscope and High-Power Laser Settings. Pediatr. Rep. 2024 , 16 , 806-815. https://doi.org/10.3390/pediatric16030068

Tatanis V, Spinos T, Lamprinou Z, Kanna E, Mulita F, Peteinaris A, Achilleos O, Skondras I, Liatsikos E, Kallidonis P. Successful Treatment of Multiple Large Intrarenal Stones in a 2-Year-Old Boy Using a Single-Use Flexible Ureteroscope and High-Power Laser Settings. Pediatric Reports . 2024; 16(3):806-815. https://doi.org/10.3390/pediatric16030068

Tatanis, Vasileios, Theodoros Spinos, Zoi Lamprinou, Elisavet Kanna, Francesk Mulita, Angelis Peteinaris, Orthodoxos Achilleos, Ioannis Skondras, Evangelos Liatsikos, and Panagiotis Kallidonis. 2024. "Successful Treatment of Multiple Large Intrarenal Stones in a 2-Year-Old Boy Using a Single-Use Flexible Ureteroscope and High-Power Laser Settings" Pediatric Reports 16, no. 3: 806-815. https://doi.org/10.3390/pediatric16030068

Article Metrics

Article access statistics, further information, mdpi initiatives, follow mdpi.

Subscribe to receive issue release notifications and newsletters from MDPI journals

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings
• My Bibliography
• Collections
• Citation manager

Save citation to file

Email citation, add to collections.

• Create a new collection
• Add to an existing collection

Add to My Bibliography

Your saved search, create a file for external citation management software, your rss feed.

• Search in PubMed
• Search in NLM Catalog
• Add to Search

Alcohol Intake and Prevalent Kidney Stone: The National Health and Nutrition Examination Survey 2007-2018

Affiliations.

• 1 Division of Kidney Diseases and Hypertension, Alpert Medical School of Brown University, Providence, RI 02903, USA.
• 2 Lifespan Biostatistics, Epidemiology, Research Design, and Informatics Core, Providence, RI 02903, USA.
• 3 Division of Kidney Diseases and Hypertension, University of Colorado School of Medicine, Aurora, CO 80045, USA.
• 4 Division of Kidney Diseases and Hypertension, Department of Medicine, Brown Physicians Inc., 375 Wampanoag Trail, East Providence, RI 02915, USA.
• PMID: 39275244
• PMCID: PMC11397207
• DOI: 10.3390/nu16172928

The association of alcohol intake with kidney stone disease (KSD) is not clear based on current clinical evidence. We examined the National Health and Nutrition Examination Survey (NHANES) 2007-2018 and used logistic regression analyses to determine the independent association between alcohol intake and prevalent KSD. In total, 29,684 participants were eligible for the final analysis, including 2840 prevalent stone formers (SFs). The mean alcohol intake was 37.0 ± 2.4 g/day among SFs compared to 42.7 ± 0.9 among non-SFs ( p = 0.04). Beer [odds ratio (OR) = 0.76, 95% CI: 0.61-0.94, p = 0.01] and wine (OR = 0.75, 95% CI: 0.59-0.96, p = 0.03) intakes were strongly associated with lower odds of prevalent KSD, while liquor intake had no association. Furthermore, the effects of beer and wine intakes on stone formation were dose-dependent. The OR for comparing participants drinking 1-14 g/day of beer to non-drinkers was 1.41 (95%CI: 0.97-2.05, p = 0.07), that of >14-≤28 g/day of beer to non-drinkers was 0.65 (95% CI: 0.42-1.00, p = 0.05), that of >28-≤56 g/day of beer to non-drinkers was 0.60 (95% CI: 0.39-0.93, p = 0.02), and that of >56 g/day of beer to non-drinkers was 0.34 (95% CI: 0.20-0.57, p < 0.001). Interestingly, the effect of wine intake was only significant among participants drinking moderate amounts (>14-28 g/day), with an OR of 0.54 (95% CI: 0.36-0.81, p = 0.003) compared to non-drinkers, but this effect was lost when comparing low-level (1-14 g/day) and heavy (>28 g/day) wine drinkers to non-drinkers. These effects were consistent in spline models. This study suggests that both moderate to heavy beer intake and moderate wine intake are associated with a reduced risk of KSD. Future prospective studies are needed to clarify the causal relationship.

Keywords: alcohol intake; beer; kidney stone; liquor; wine.

PubMed Disclaimer

Conflict of interest statement

The authors declare that they have no competing interests.

Selection of study population.

Odds ratios of prevalent kidney…

Odds ratios of prevalent kidney stone by restricted cubic splines for exclusive beer…

Odds ratios of prevalent kidney stone by restricted cubic splines for exclusive wine…

Odds ratios of prevalent kidney stone for continuous liquor intake among current drinkers…

• Scales C.D., Jr., Smith A.C., Hanley J.M., Saigal C.S. Prevalence of kidney stones in the United States. Eur. Urol. 2012;62:160–165. doi: 10.1016/j.eururo.2012.03.052. - DOI - PMC - PubMed
• Saigal C.S., Joyce G., Timilsina A.R. Direct and indirect costs of nephrolithiasis in an employed population: Opportunity for disease management? Kidney Int. 2005;68:1808–1814. doi: 10.1111/j.1523-1755.2005.00599.x. - DOI - PubMed
• Aruga S., Honma Y. Renal calcium excretion and urolithiasis. Clin. Calcium. 2011;21:1465–1472. - PubMed
• Borghi L., Meschi T., Amato F., Briganti A., Novarini A., Giannini A. Urinary volume, water and recurrences in idiopathic calcium nephrolithiasis: A 5-year randomized prospective study. J. Urol. 1996;155:839–843. doi: 10.1016/S0022-5347(01)66321-3. - DOI - PubMed
• Curhan G.C., Willett W.C., Rimm E.B., Spiegelman D., Stampfer M.J. Prospective study of beverage use and the risk of kidney stones. Am. J. Epidemiol. 1996;143:240–247. doi: 10.1093/oxfordjournals.aje.a008734. - DOI - PubMed
• Search in MeSH

Grants and funding

• N/A/Foundation grant from Brown Physicians Inc

LinkOut - more resources

Full text sources.

• PubMed Central
• MedlinePlus Health Information

• Citation Manager

NCBI Literature Resources

MeSH PMC Bookshelf Disclaimer

The PubMed wordmark and PubMed logo are registered trademarks of the U.S. Department of Health and Human Services (HHS). Unauthorized use of these marks is strictly prohibited.

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 18 September 2024

Integrating neural networks with advanced optimization techniques for accurate kidney disease diagnosis

• Samar Elbedwehy 1 ,
• Esraa Hassan 2 ,
• Abeer Saber 3 &
• Rady Elmonier 4  

Scientific Reports volume  14 , Article number:  21740 ( 2024 ) Cite this article

Metrics details

• Kidney diseases
• Radiography

Kidney diseases pose a significant global health challenge, requiring precise diagnostic tools to improve patient outcomes. This study addresses this need by investigating three main categories of renal diseases: kidney stones, cysts, and tumors. Utilizing a comprehensive dataset of 12,446 CT whole abdomen and urogram images, this study developed an advanced AI-driven diagnostic system specifically tailored for kidney disease classification. The innovative approach of this study combines the strengths of traditional convolutional neural network architecture (AlexNet) with modern advancements in ConvNeXt architectures. By integrating AlexNet’s robust feature extraction capabilities with ConvNeXt’s advanced attention mechanisms, the paper achieved an exceptional classification accuracy of 99.85%. A key advancement in this study’s methodology lies in the strategic amalgamation of features from both networks. This paper concatenated hierarchical spatial information and incorporated self-attention mechanisms to enhance classification performance. Furthermore, the study introduced a custom optimization technique inspired by the Adam optimizer, which dynamically adjusts the step size based on gradient norms. This tailored optimizer facilitated faster convergence and more effective weight updates, imporving model performance. The model of this study demonstrated outstanding performance across various metrics, with an average precision of 99.89%, recall of 99.95%, and specificity of 99.83%. These results highlight the efficacy of the hybrid architecture and optimization strategy in accurately diagnosing kidney diseases. Additionally, the methodology of this paper emphasizes interpretability and explainability, which are crucial for the clinical deployment of deep learning models.

Introduction

Kidney diseases have emerged as a significant global health concern, with chronic kidney disease affecting over 10% of the world's population. This condition, predicted to rise to the fifth leading cause of death by 2040, underscores the pressing need for effective control measures. Among the prevalent kidney ailments impeding normal renal function, kidney cysts, nephrolithiasis (kidney stones), and renal cell carcinoma (kidney tumor) pose substantial threats 1 . Kidney cysts, fluid-filled pockets on the kidney's surface, and nephrolithiasis, involving crystal concretion formation, impact approximately 12% of the global population. Renal cell carcinoma is identified as one of the top ten most common cancers worldwide. There are different types of data that researchers handled it; Text and images. Text datasets Also often contain valuable information derived from medical records, pathology reports, and patient histories, which can be leveraged to train machine learning models 2 , 3 , 4 , 5 .

Also, diagnostic tools such as X-ray, computed tomography (CT), B-ultrasound, and magnetic resonance imaging (MRI) play crucial roles in conjunction with pathology tests for accurate kidney disease diagnosis. CT scans, particularly valuable for their three-dimensional insights and detailed slice-by-slice imaging, offer a comprehensive understanding of kidney anatomy 6 .

Recognizing the urgency of addressing these challenges, the advancement of deep learning in vision tasks presents a compelling opportunity. Building artificial intelligence (AI) models capable of efficiently detecting kidney radiological findings has become imperative to assist medical professionals and alleviate the suffering of individuals affected by kidney diseases. While some studies have explored this domain, the scarcity of publicly available datasets remains a hindrance. Furthermore, past research has often relied on traditional machine learning algorithms, focusing on the classification of single disease classes, such as cysts, tumors, or stones, and occasionally utilizing ultrasound images. In light of these considerations, there is a growing need to expand the scope of AI applications, leveraging deep learning advancements for a more comprehensive approach to kidney disease detection.

• Feature concatenation

Feature concatenation plays a crucial role in enhancing the effectiveness of deep learning models, especially in tasks such as image classification. By combining different types of features extracted from diverse sources, feature concatenation enables the creation of a more comprehensive and informative representation of the input data. This process allows the model to leverage complementary information embedded in various aspects of the data, such as color, texture, or spatial features. Unlike traditional single-feature approaches, feature concatenation enables the model to capture a richer set of characteristics, potentially improving its ability to generalize and make accurate predictions. Moreover, this technique facilitates the integration of information from different modalities or feature extraction methods, leading to a more robust and nuanced representation. In essence, feature concatenation serves as a powerful tool for refining the input representation, contributing to the model's overall performance and its capacity to handle complex patterns and relationships within the data.

The main contributions of this study are as follows

Novel classification method: the paper proposes a new approach for classifying kidney diseases that demonstrates robust performance across various datasets, emphasizing the importance of interpretability and explainability for clinical applications.

Advanced integration of neural networks: this study integrates features from AlexNet and ConvNexT to create a comprehensive and informative feature representation. This fusion leverages the strengths of both architectures, resulting in superior performance compared to individual models.

Enhanced model performance: By combining AlexNet and ViT, the paper achieved improved discriminative ability, capturing a broader range of visual features and surpassing the performance of the individual models.

Optimized training process: this study introduced a custom optimization technique based on Adam that dynamically adjusts the step size according to the gradient norm, leading to more efficient convergence in training the merged AlexNet and ConvNexT models.

The rest of the paper is organized as follows; in the next section; literature reviews. In Sect. " Motivation ", the motivation. In Sect. " Proposed methodology ", the proposed methodology is used in this paper, followed by Sect. “ Experiments and results ”. Finally, in Sect. " Conclusions "; the paper is concluded with future work.

Literature reviews

The classification of kidney diseases is a pivotal area of research that holds significant implications for clinical diagnosis, treatment planning, and patient management. As the understanding of renal disorders continues to evolve, there has been a growing body of literature dedicated to exploring various methodologies and techniques for accurate and efficient kidney disease classification. This literature review seeks to provide a comprehensive overview of the existing research landscape, delving into the diverse approaches employed in the classification of kidney conditions. From traditional methods to the latest advancements in machine learning and deep learning, this review aims to distill key insights and trends, shedding light on the progress made in enhancing diagnostic accuracy and paving the way for more effective therapeutic interventions. Through a systematic exploration of relevant studies, this literature review endeavors to offer a synthesis of knowledge that not only underscores the current state of kidney disease classification but also identifies potential avenues for future research and technological innovation in this critical domain. Parakh et al. 7 proposed the initial convolutional neural network (CNN) was responsible for delineating the urinary tract's extent, while the second CNN focused on identifying the presence of stones. The authors created nine model variations by combining different training data sources (S1, S2, or both, denoted as SB) with pre-trained CNNs using ImageNet and GrayNet, as well as without pretraining (Random). The accuracy of GrayNet-SB, at 95%, surpassed that of ImageNet-SB (91%) and Random-SB (88%).

The research of Kuo et al. 8 aims to enhance the prediction of kidney function and chronic kidney disease (CKD) through kidney ultrasound imaging, develop a model integrating the ResNet architecture, pre-trained on the ImageNet dataset, to estimate the glomerular filtration rate (eGFR) and CKD status from 4505 labeled kidney ultrasound images. The model demonstrated a strong correlation (Pearson coefficient of 0.741) between AI-based and creatinine-based GFR estimations and achieved 85.6% accuracy in classifying CKD status, outperforming experienced nephrologists (60.3%–80.1%).

Sudharson et al. 9 utilized an ensemble technique, amalgamating diverse pre-trained Deep Neural Networks (DNNs) such as ResNet-101, ShuffleNet, and MobileNet-v2. The ultimate predictions were determined through the majority voting technique, resulting in a peak classification accuracy of 96.54% during testing with high-quality images and 95.58% during testing with noisy images.

Aksakallı et al. 10 proposed the examination encompassed diverse machine learning approaches, including Decision Trees (DT), Random Forest (RF), Support Vector Machines (SVC), Multilayer Perceptron (MLP), K-Nearest Neighbor (kNN), Naive Bayes (BernoulliNB), and deep neural networks employing Convolutional Neural Network (CNN). The experimental outcomes revealed that the Decision Tree Classifier (DT) yielded the most favorable classification results. Specifically, this method attained the highest F1 score, achieving a success rate of 85.3% when employing the S + U sampling method.

Liu et al. 11 focuse on making deep learning techniques more accessible for clinical users in the field of microscopic image classification by developing AIMIC, out-of-the-box software that requires no programming knowledge. AIMIC integrates advanced deep learning methods and data preprocessing techniques, allowing users to train new networks and infer unseen samples seamlessly. The platform was evaluated on four benchmark microscopy image datasets, demonstrating its effectiveness in selecting suitable algorithms for entry-level practitioners. Notably, the ResNeXt-50–32 × 4d model achieved the highest performance with an average accuracy of 96.83% and an average F1-score of 96.82%, making it the preferred choice for microscopic image classification. Additionally, MobileNet-V2 provided a good balance between accuracy (95.72%) and computational cost, with an inference time of 0.109 s per sample, making it a viable option for scenarios with limited computing resources.

Srivastava et al. 12 used machine learning models (SVM, KNN, Random Forest, Decision Tree, AdaBoost) with the normalized dataset with an accuracy of 98.75%. Baygin et al. 13 proposed a novel transfer learning-based image classification method called ExDark19. This method utilized iterative neighborhood component analysis (INCA) to select the most informative feature vectors, which were then input into a k nearest neighbor (kNN) classifier for kidney stone detection. Their results achieved an accuracy of 99.22% with a ten-fold cross-validation strategy and 99.71% using the hold-out validation method.

Nazmul Islam et al. 14 employed a total of six machine learning models, with three being founded on advanced variants of Vision Transformers, namely EANet, CCT, and Swin Transformers. The remaining three models were based on deep learning architectures, ResNet, VGG16, and Inception v3, with adjustments made to their final layers. Despite commendable performances from the VGG16 and CCT models, the Swin Transformer emerged as the top performer in terms of accuracy, achieving an impressive accuracy rate of 99.30 percent. In this investigation, diverse physiological parameters were considered alongside the application of various machine learning (ML) techniques. Different ML models, including Support Vector Machines (SVM), K-Nearest Neighbors (KNN), Random Forest, Decision Tree, and AdaBoost, were trained using a normalized dataset, resulting in an impressive accuracy of 98.75%, perfect sensitivity (100%), high specificity of 96.55%, and a notable f1 score of 99.03%.

Subedi et al. 15 explore the potential of a novel model called Vision Transformer (ViT), which was initially designed for natural language processing (NLP) tasks but shows promise for medical image classification. ViT’s capabilities are further enhanced by coupling it with Fully Connected Networks (FCN). This combination merges the feature extraction capabilities of ViT with the classification ability of FCN, ultimately overcoming the challenge of detecting kidney-related issues with greater accuracy and reliability with an accuracy of 99.64%.

Asif et al. 16 introduced "StoneNet” which is based on MobileNet using depthwise separable convolution, offering a low-cost solution compared to existing models with drawbacks such as high computational costs and lengthy training times. Their model achieved accuracy at 97.98%, with short training and testing times of 996.88 s and 14.62 s, respectively. Qadir et al. 17 focused on the Densenet-201 model for feature extraction with Random Forest being the chosen method. They achieved an accuracy rate of 99.719%. Table 1 presents the related work for the kidney classification.

Sasikaladevi et al. 18 address the critical need for early and automatic detection of chronic kidney disease (CKD) from radiology images using deep learning techniques. The dataset used contains 12,446 unique CT scan images. Deep features were extracted from these images, and hyperedges were generated to construct hypergraphs representing the renal images. These hypergraphs were then used in a hypergraph convolutional neural network for representational learning. The model was validated using a hold-out dataset, and deep learning metrics including precision, recall, accuracy, and F1 score were used to evaluate its performance. The proposed model demonstrated a superior validation accuracy of 99.71%, outperforming other state-of-the-art algorithms. This robust digital-twin model facilitates early diagnosis of kidney diseases and aids nephrologists in better prognosis of kidney-related abnormalities.

The urgent need to improve patient care and medical diagnostics in the field of renal health is the driving force behind the kidney classification paper. Kidney illnesses are a major global health concern, encompassing both acute and chronic ailments. Timely and accurate categorization of these ailments is essential for efficient treatment strategy development and patient supervision.

Several factors contribute to the motivation for kidney classification research:

Clinical Importance: Diagnosing kidney disorders accurately can be challenging due to their wide range of etiologies and symptoms. Enhancing classification techniques helps medical professionals better comprehend various kidney disorders and customize treatment plans based on individual disease profiles.

Early Identification and Intervention: It's critical to identify kidney disorders early to launch prompt interventions that can halt the disease's progression and enhance patient outcomes. Classification models can help detect kidney function issues early on, which can result in more proactive and focused medical interventions.

Application of Advanced Technologies: The development of complex models for the classification of renal disease is made possible by advances in machine learning, deep learning, and image processing techniques. Making use of these technologies has the potential to completely transform how accurate and effective diagnostic procedures are.

Proposed methodology

The paper discusses the impact of the concatenating features for enhancing the accuracy of kidney disease classification using the merging of Alex-Net 19 with other models such as (ViT 20 , Swin 21 , and ConvNexT 22 ) and also the impact with using the modified Adam optimizer “Custom-Adam” instead of the popular optimizer “Adam”.

The paper compared its performance with more recent architectures such as VGG and ResNet. The results show that the pre-trained VGG and ResNet models achieved accuracies of 91.73% and 94.63%, respectively. In contrast, more advanced models such as Vision Transformer (ViT), Swin Transformer, and ConvNexT achieved higher accuracies of 98.71%, 96.44%, and 96.44%, respectively. These findings highlight the superior performance of these newer architectures over Alex-Net. While Alex-Net has a well-established reputation in image classification tasks as its architecture is known for efficient feature extraction, which is crucial for accurately classifying kidney diseases from medical images.

Transformer models which include ViT and Swin have demonstrated remarkable performance in various computer vision tasks, particularly in capturing long-range dependencies and spatial relationships within images. For example, the main purpose for using the ViT model is self-attention mechanism allows it to capture global contextual information in images, enabling it to identify complex patterns and long-range relationships. But Swin optimizes the attention computation in Vision Transformers by limiting self-attention to non-overlapping local windows. This shifted window approach reduces the normally quadratic complexity of ViT to linear complexity concerning image size, making Swin more computationally efficient. Also, Swin is a hierarchical vision transformer that progressively merges adjacent patches as the network deepens. This hierarchical structure enables the model to manage features at various scales, enhancing the learning of robust and discriminative features compared to convolutional neural networks. But with ConvNexT model, incorporates modern techniques like hierarchical design and larger kernel sizes, enhancing its ability to handle diverse image features while maintaining the simplicity of traditional CNNs. The paper included these models to explore their potential to extract relevant features from medical images, which could contribute to improving diagnostic accuracy.

On the other side, changing the optimizer can significantly impact model accuracy, convergence speed, generalization ability, and overall stability. Therefore, choosing the right optimizer is crucial for optimizing machine learning models. The paper compared the effect of Adam 23 and Custom_Adam optimizer on the dataset to find the Custom_Adam is better in most cases while the primary difference between the standard Adam optimizer and the Custom_Adam lies in the additional calculation and utilization of the gradient norm in the custom version. Specifically, Custom_Adam computes the norm of the gradient (denoted as norm_value) for each parameter \(\theta\) with a non-None gradient:

This norm is then used in the custom update rule. The _update_rule method in Custom_Adams incorporates this norm_value along with the parameter \(\theta\) , gradient \({g}_{t}\) ​, and state during the update process, which can be expressed as:

The parameter update in the standard Adam is as:

Additionally, Custom_Adam overrides the step method to include the gradient norm calculation and the call to the _update_rule, whereas the standard Adam optimizer utilizes its default step method without these extra computations. This enhancement allows Custom_Adam to adapt the learning rate based on the gradient's scale, potentially improving optimization performance. See the algorithm as the following .

Algorithm: Custom_Adam

To accomplish this, the two actions listed can be taken: first; compare using four single vision models (ViT, Alex-Net, Swin, and ConvNexT) for extracting the features from images by using the optimizer Adam and the Custom_Adam. The second is to improve the extracting feature process using the concatenating features from the four vision models with the best optimizer that got from the first action; the vision models are (“Swin + ConvNexT”, “Alex-Net + ViT”, “Alex-Net + Swin” and “Alex-Net + ConvNexT” ). The paper finds as in Fig.  3 that concatenating the models Alex-Net with ConvNexT with Custom_Adam optimizer is the best value in accuracy 99.85% with metrics used for the evaluation such as average precision, recall, and specificity, reaching 99.89%, 99.95%, and 99.83% respectively.

The methodology of this study for kidney classification involves several steps as in Fig.  1

The methodology of this study for Kidney classification.

Image loading from the directory then applied using T.Compose to augment the training data, these transformations include random horizontal and vertical flips, random color jitter, resizing to 256 * 256 pixels, center cropping to 224 * 224 pixels, conversion to a PyTorch tensor, normalization using ImageNet mean and standard deviation, and random erasing with a probability of 0.1.

Load pre-trained models (AlexNet and ConvNexT) then freeze the parameters of the loaded models and create a new model by concatenating the output features of the two models and then adding a classifier layer.

Define a custom optimizer class that inherits from Adam with the modifications.

Define functions to get data loaders for training and validation then implement data loading and augmentation for the training set and the validation set.

Define the training loop using the optimizer and the loss “CrossEntropyLoss”.

Evaluate the model using the confusion matrix and the learning curve for the loss and the accuracy.

The paper used the dataset that originated from various hospitals in Dhaka, Bangladesh, where patients had previously received diagnoses related to kidney tumors, cysts, normal conditions, or stone findings. The gathered data from the Picture Archiving and Communication System (PACS), incorporating both Coronal and Axial cuts from contrast and non-contrast studies covering the entire abdomen and urogram. Subsequently, patient information and metadata were excluded from the Dicom images, and the images were converted to a lossless jpg format. To ensure accuracy, each image finding underwent verification by both a radiologist and a medical technologist after the conversion process 14 . The dataset contains 12,446 unique data within it which the cyst contains 3709, normal 5077, stone 1377, and tumor 2283. As shown in Fig 2 . The sample of the dataset used.

Sample images from the dataset.

Experiments and results

This study used the assembled and annotated 12,446 CT 14 whole abdomen and urogram images that contained four classes Cyst, Normal, Stone, and Tumor as in Fig.  2 . The paper divided the dataset into training and validation using augmentation to overcome the overfitting problem such as RandomHorizontalFlip, RandomVerticalFlip, CenterCrop, and Normalize the images. After augmentation training dataset be 19,450 instead of 9725 and the validation be 5442 instead of 2721.

The hyperparameter settings for the best model (Alex-NeT + ConvNexT with custom-Adam optimizer) are as follows: learning rate with 1e-4, Epochs = 100, loss = CrossEntropyLoss, Optimizer = custom-Adam and batch_size = 32. The paper trained the models using pytorch with a laptop with one GPU (2060 RTX). Figures  3 and 4 show the (training and validation loss) and (training and validation accuracy) respectively while Fig.  5 shows the Precision, Recall, and F1-score for the model.

Training and Validation Loss.

Training and Validation Accuracy.

Precision, Recall, and F1-score for Alex-NeT + ConvNexT with custom-Adam optimizer.

Performance evaluation methods

The evaluation of the eight models involves an analysis based on parameters such as accuracy in training, sensitivity (or recall), and precision (or positive predictive value - PPV). To calculate precision, and Recall, the paper utilizes true positive (TP), false positive (FP), true negative (TN), and false negative (FN) samples. Recall, also known as sensitivity, is determined by dividing the number of true positives by the sum of true positives and false negatives. In medical diagnosis, high recall is imperative for accurately identifying individuals with the disease, as overlooking the positive category can result in serious consequences like misdiagnosis and treatment delays. Precision (PPV) becomes crucial when assessing the proportion of predicted positive examples that are genuinely positive. Precision is calculated by dividing the number of true positives by the sum of true positives and false positives. In the realm of medical imaging, achieving high precision is highly desirable. The F1 score for all models is derived from the sensitivity and precision values. The provided formulas are applied to calculate accuracy, precision, sensitivity, and the F1 score 24 .

where, i=class of the kidney (Cyst or Normal or Stone or Tumor), TP= True Positive, FN= False Negative, TN=True Negative.

Table 2 shows the comparison between single vision models using the Adam optimizer and custom_Adam optimizer for the four classes of kidney diseases with some factors such as; accuracy, precision, recall, f-score, and the average for the four classes.

The presented table summarizes the performance of various models, each employing different optimizers, in distinguishing between four classes: Cyst, Normal, Stone, and Tumor. Notably, Vision Transformer (ViT) models, both with Adam and Custom_Adam optimizers, consistently demonstrate robust accuracy, precision, and recall across the specified classes, showcasing their effectiveness in image classification tasks. Swin and ConvNexT models also exhibit commendable performance, with high accuracy and stable precision-recall metrics. Alex-Net models, while slightly lagging in accuracy, still demonstrate competitive results. The ViT model with Adam optimizer consistently demonstrates high accuracy across all classes, making it a strong contender. Precision and recall are often critical in medical imaging; the balance between the two might be preferred.

Here, the study presents the best confusion matrix for the four individual vision models utilizing Adam and custom_Adam, which demonstrates improved results in Figs. 6 , 7 , 8 , and 9 .

ViT with Adam optimizer model.

Alex-Net with custom_Adam optimizer.

Swin with custom_Adam optimizer.

ConvNexT with Adam optimizer.

Visualizing results using class-wise error rates is also essential for the evaluation of image classification models. This approach provides a detailed view of the model's performance across different categories. Unlike overall accuracy metrics, which aggregate performance across all classes, class-wise error rates highlight disparities in classification performance. It can offer a comprehensive understanding of model efficiency. Here is the class-wise error rate for the best four models the paper used in Figs. 10 , 11 , 12 , and 13 .

Class-wise error rate for ViT with Adam optimizer model.

Class-wise error rate for Alex-Net with custom_Adam optimizer.

Class-wise error rate for Swin with custom_Adam optimizer.

Class-wise error rate for ConvNexT with Adam optimizer.

The summarized comparison of the class-wise error rate between the best four models in Fig.  14

Class-wise error rate for the best four models.

As in Fig.  14 , all models consistently achieve near-perfect performance, with the second model (Swin with custom_Adam optimizer) achieving perfect classification. The error rates vary, with the third model (Alex-Net with custom_Adam optimizer) showing higher error rates, while the final model (ViT with Adam optimizer) shows the best performance. All models demonstrate strong performance with low error rates, with the second and fourth models showing the best performance. The best overall model appears to be the "ViT with Adam optimizer model", as it achieves the lowest error rates across most classes, demonstrating consistent and strong performance in classifying 'Cyst', 'Normal', 'Stone', and 'Tumor' samples.

Table 3 shows the comparison between concatenated vision models using Adam and custom_Adam optimizer for the four classes of kidney diseases with some factors such as; accuracy, precision, recall, f-score, and the average for the four classes.

Table 3 presents the effect of the concatenated features between the models. Alex-Net + ConvNexT with the custom_Adam stand out with the highest accuracy of 99.85%. On the other hand, the model with the lowest accuracy among those provided, Swin + ConvNexT with the custom_Adam optimizer with an accuracy of 98.75% has the lowest accuracy but its balanced precision and recall suggest effectiveness across various classes. But Alex-Net + ConvNext with the custom_Adam stands out with consistently high average precision (0.9989) and recall (0.9995) values, indicating robust performance across all classes. Among the provided models, the custom_Adam optimizer consistently outperforms the standard Adam optimizer in terms of accuracy, precision, recall, and F1-score in all concatenated models specifically the Alex-Net model with any Transformer model with the dynamic adjustment of the step size based on the norm of the gradient except of the Swin + ConvNexT model which give the less result with the custom_Adam and the Adam optimizer which may because the different architectures that make the model more complexity. Also if the gradient flow between Swin and ConvNexT is not well-aligned, the gradients might not propagate effectively during training, leading to convergence challenges.

Here, the study presents the best confusion matrix for the four concatenated vision models utilizing Adam and custom_Adam, which demonstrates the best results in Figs. 15 , 16 , 17 , and 18 .

Swin + ConvNexT with Adam optimizer.

Alex-Net + Swin with custom_Adam optimizer.

Alex-Net + ViT with custom_Adam optimizer.

Alex-Net + ConvNexT with custom_Adam optimizer.

Here is the class-wise error rate of the best concatenated models in Figs. 19 , 20 , 21 , and 22 .

Class-wise error rate for Swin + ConvNexT with Adam optimizer.

Class-wise error rate for Alex-Net + Swin with custom_Adam optimizer.

Class-wise error rate for Alex-Net + ViT with custom_Adam optimizer.

Class-wise error rate for Alex-Net + ConvNexT with custom_Adam optimizer.

The summarized comparison for the class-wise error rate between the best four concatenated models in Fig.  23

Class-wise error rate for the best four concatenated models.

As in Fig.  23 , all models perform well in classifying 'Cyst' samples, with the final model (Alex-Net + ConvNexT with Adam Optimizer) showing perfect performance. Also, all models consistently achieve near-perfect or perfect performance in classifying 'Normal' samples, with multiple models achieving perfect performance. The error rates vary slightly, but all models generally perform well, with the second model (Alex-Net + Swin with custom_Adam Optimizer) showing the best improvement. All models demonstrate strong performance, with very low error rates across the board. The second model (Alex-Net + Swin with custom_Adam Optimizer) and the final model (Alex-Net + ConvNexT with Adam Optimizer) show perfect or near-perfect performance. The best overall model appears to be the "Alex-Net + ConvNexT with Adam Optimizer Model", as it achieves perfect classification in the 'Cyst' and 'Normal' classes, very low error rates in the 'Stone' class, and almost perfect performance in the 'Tumor' class. This model consistently demonstrates strong performance across all classes, making it the most reliable and effective model in this comparison.

No. parameters of different models

One essential feature that greatly affects a neural network model's capacity, efficiency, and flexibility is the number of parameters. Deep learning models consist of several layers, each of which has weights and biases that add to the total number of parameters. Greater representational capacity is often possessed by larger, more parameterized models, which allows them to learn complex characteristics and relationships in data. Conversely, more compact models with fewer parameters could be less prone to overfitting and more computationally efficient, which makes them appropriate for jobs requiring sparse data. As shown in Table 4 , the total number of parameters and trainable parameters for the single models and the concatenated models used in this paper. It's generally more meaningful to focus on "Trainable parameters" rather than "Total number of parameters." because not all parameters in a model may be trainable, as some might be fixed or non-trainable. As in Table 4 , the model with the least parameters is Swin, and the model with the most parameters is Alex-Net + ConvNexT. Larger parameter counts are often associated with better model accuracy, so the progression from the model with the least parameters to the most parameters could represent an increase in model capacity and, potentially, accuracy as in Fig. 24 .

Trainable parameters for the models used in the paper.

Time evaluation

For each of the 8 tested single models, the study compared the time taken for training for each model to get the less time that was taken. As shown in Fig. 25 , Alex-Net with Adam optimizer was the fastest in training as it took the least training time (50 minutes) with an accuracy of 96.32 followed by Swin with Adam optimizer which took 59 minutes with an accuracy of 96.44 then Alex-Net+ custom_Adam that took 66 minutes with accuracy 96.91 while the best one which is ViT with Adam optimizer took 5 hours approximately with accuracy 98.71 while the ConvNexT with custom_Adam optimizer took the longest time with around 10 hours with an accuracy of 96.62.

Time evaluation for training the eight single models.

As shown in Fig.  26 , the concatenated models for each of the 8 tested concatenated models, the paper also compared the time taken for training for each model to get the less time that taken., Swin + Alex-Net with custom_Adam optimizer was the fastest in training as it took the least training time (2 h and 30 min) with an accuracy of 99.78 followed by Swin + ConvNexT with Adam optimizer which took around 3 h and a half with an accuracy of 99.12 while ConvNexT + Alex-Net, Swin + Alex-Net with Adam optimizer and Swin + ConvNexT with custom_Adam optimizer took the same time around 4 h and a half with accuracies 99.63, 99.45 and 98.75 respectively. The best one which is ConvNexT + Alex-Net with cuatom_Adam optimizer took 6 h approximately with an accuracy of 99.85. While the Alex-Net + ViT with custom_Adam optimizer took the longest time around 7 h with an accuracy of 99.74.

Time evaluation for training the eight concatenated models.

Conclusions

This study explored the impact of feature concatenation and optimizer selection on neural network performance. The experimental results reveal that concatenating features, such as Alex-Net + ConvNexT, in combination with the custom_Adam optimizer, achieved an impressive accuracy of 99.85%. This highlights the benefits of integrating diverse model architectures and optimizing strategies to capture complex patterns and correlations in data. The custom_Adam optimizer demonstrated superior performance compared to the standard Adam optimizer across all concatenated models, excelling in accuracy, precision, recall, and F1-score. Particularly notable was its effect when paired with Transformer models, where dynamic step size adjustments based on gradient norms contributed to consistently high average recall and accuracy. The trade-off between model capacity and efficiency was evident, with the Swin model, despite its fewer parameters, performing competitively. This underscores its utility in scenarios where computational efficiency and reduced overfitting are critical. While larger models like Alex-Net + ConvNexT exhibited higher accuracy, the Swin + Alex-Net combination offered a balanced approach with a training duration of 2 h and 30 min and an accuracy of 99.78%. Conversely, the Alex-Net + ViT configuration, though achieving 99.74% accuracy, required the longest training time of approximately 7 h.

Data availability

The data that support the findings of this study are available from https://www.kaggle.com/datasets/nazmul0087/ct-kidney-dataset-normal-cyst-tumor-and-stone .

Foreman, K. J. et al. Forecasting life expectancy, years of life lost and all-cause and cause-specific mortality for 250 causes of death: Reference and alternative scenarios for 2016–40 for 195 countries and territories. Lancet 392 , 2052–2090. https://doi.org/10.1016/S0140-6736(18)31694-5.[PMCfreearticle][PubMed][CrossRef][GoogleScholar] (2018).

Article   PubMed   PubMed Central   Google Scholar  

Jain, D. & Singh, V. A novel hybrid approach for chronic disease classification. Int. J. Healthcare Inf. Syst. Informat. (IJHISI) 15 (1), 1–19 (2020).

Article   Google Scholar  

Jain, D. & Singh, V. A two-phase hybrid approach using feature selection and adaptive SVM for chronic disease classification. Int. J. Comput. Appl. 43 (6), 524–536 (2021).

Google Scholar  

Singh, V., Asari, V. K. & Rajasekaran, R. A deep neural network for early detection and prediction of chronic kidney disease. Diagnostics 12 (1), 116 (2022).

Singh, V., & Jain, D. A hybrid parallel classification model for the diagnosis of chronic kidney disease. (2023).

Saw, K. C. et al. Helical CT of urinary calculi: Effect of stone composition, stone size, and scan collimation. Am. J. Roentgenol. 175 (2), 329–332 (2000).

Parakh, A. et al. Urinary stone detection on CT images using deep convolutional neural networks: Evaluation of model performance and generalization. Radiol. Artif. Intell. 1 (4), 180066 (2019).

Kuo, C.-C. et al. Automation of the kidney function prediction and classification through ultrasound-based kidney imaging using deep learning. NPJ Digital Med. 2 (1), 29 (2019).

Sudharson, S. & Kokil, P. An ensemble of deep neural networks for kidney ultrasound image classification. Comput. Methods Progr. Biomed. 197 , 105709 (2020).

Aksakalli, I., Kaçdioğlu, S. & Hanay, Y. S. Kidney x-ray images classification using machine learning and deep-learning methods. Balkan J. Electr. Comput. Eng. 9 (2), 44–551 (2021).

Liu, R. et al. AIMIC: Deep learning for microscopic image classification. Comput. Methods Program. Biomed. 226 , 107162 (2022).

Srivastava, S., Yadav, R.K., Narayan, V., & Mall, P.K. An ensemble learning approach for chronic kidney disease classification. J. Pharmaceut. Negative Results , 2401–2409 (2022).

Baygin, M. et al. Exemplar Darknet19 feature generation technique for automated kidney stone detection with coronal CT images. Artif. Intell. Med. 127 , 102274 (2022).

Article   PubMed   Google Scholar  

Islam, M. N. et al. Vision transformer and explainable transfer learning models for auto detection of kidney cyst, stone and tumor from CT-radiography. Sci. Rep. 12 (1), 1–14 (2022).

Subedi, R., Timilsina, S. & Adhikari, S. Kidney CT scan image classification using modified vision transformer. J. Eng. Sci. 2 (1), 24–29 (2023).

Asif, S., Zhao, M., Chen, X. & Zhu, Y. StoneNet: An efficient lightweight model based on depthwise separable convolutions for kidney stone detection from CT images. Interdiscipl. Sci. Comput. Life Sci. 15 (4), 633–652 (2023).

Qadir, A. M. & Abd, D. F. Kidney diseases classification using hybrid transfer-learning densenet201-based and random forest classifier. Kurdistan J. Appl. Res. 7 (2), 131–144 (2022).

Sasikaladevi, N. & Revathi, A. Digital twin of renal system with CT-radiography for the early diagnosis of chronic kidney diseases. Biomed. Signal Process. Control 88 , 105632 (2024).

Krizhevsky, A., Sutskever, I., & Hinton, G. E. Imagenet classification with deep convolutional neural networks. Adv. Neural Inf. Process. Syst. 25 (2012).

Dosovitskiy, A., Beyer, L., Kolesnikov, A., Weissenborn, D., Zhai, X., Unterthiner, T., & Dehghani, M. et al. An image is worth 16x16 words: Transformers for image recognition at scale. arXiv preprint arXiv:2010.11929 (2020).

Liu, Z., Lin, Y., Cao, Y., Hu, H., Wei, Y., Zhang, Z., Lin, S., & Guo, B. Swin transformer: Hierarchical vision transformer using shifted windows. In Proceedings of the IEEE/CVF International Conference on Computer Vision , pp. 10012–10022. 2021.

Liu, Z., Mao, H., Wu, C.-Y., Feichtenhofer, C., Darrell, T., & Xie, S. A convnet for the 2020s. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pp. 11976–11986 (2022).

Kingma, D. P., & Ba, J. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980 (2014).

Tuyet, V. T. H., Hien, N. M., Quoc, P. B., Son, N. T. & Binh, N. T. Adaptive content-based medical image retrieval based on local features extraction in shearlet domain. EAI Endorsed Trans. Context-Aware Syst. Appl. 6 (17), e3 (2019).

Download references

Open access funding provided by The Science, Technology & Innovation Funding Authority (STDF) in cooperation with The Egyptian Knowledge Bank (EKB). The authors received no specific funding for this study.

Author information

Authors and affiliations.

Department of Data Science, Faculty of Artificial Intelligence, Kafrelsheikh University, Kafrelsheikh, 33511, Egypt

Samar Elbedwehy

Faculty of Artificial Intelligence, Kafrelsheikh University, Kafrelsheikh, 33511, Egypt

Esraa Hassan

Department of Information Technology, Faculty of Computers and Artificial Intelligence, Damietta University, Damietta, 34517, Egypt

Abeer Saber

Department of Internal Medicine, Faculty of Medicine, Al-Azhar University, New Damietta, Egypt

Rady Elmonier

You can also search for this author in PubMed   Google Scholar

Contributions

S.E., E.H., A.S. and R.E. contributed to the design and implementation of the research, to the analysis of the results and to the writing of the manuscript.

Corresponding author

Correspondence to Samar Elbedwehy .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Informed consent

This article does not contain any studies with human participants or animals performed by any of the authors.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Elbedwehy, S., Hassan, E., Saber, A. et al. Integrating neural networks with advanced optimization techniques for accurate kidney disease diagnosis. Sci Rep 14 , 21740 (2024). https://doi.org/10.1038/s41598-024-71410-6

Download citation

Received : 12 April 2024

Accepted : 27 August 2024

Published : 18 September 2024

DOI : https://doi.org/10.1038/s41598-024-71410-6

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Image classification
• Vision transformer

By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.