AN EXTENSIVE REVIEW OF KIDNEY STONE SYMPTOMS, TYPES AND PREDICTION METHODS IN UP TO DATE

AN EXTENSIVE REVIEW OF KIDNEY STONE SYMPTOMS, TYPES AND PREDICTION METHODS IN UP TO DATE

B.Reuben#1, C.Narmadha*2

1. # Assistant Professor, Dept of ECE, Sri Bharathi Engineering College for Women, Kaikkurichi, Tamil Nadu, India. princeton.reuben@gmail.com

   2.* Assoc.Prof & Head, Dept of ECE, Periyar Maniammai Institute of Science & Technology, (Deemed to be University), Vallam,Thanjavur 613403, Tamil Nadu, India.

   narmadhaece@pmu.edu

   Abstract Kidney stone disease is a most harmful disease which leads to kidney failure and also kills patients. The imaging techniques like CT scans, ultrasonography, X-ray and MRI scans are mostly used for easy prediction. in Recent times, Machine Learning and Deep Learning are used for effective kidney stone prediction. In this article, the review of DL model-based kidney stone prediction is discussed with several details namely Methods used in literature, performance analysis, merits and challenges respectively. The symptoms and types of stones are presented to show the variation among each stone in a kidney. Therefore, the extensive review based on kidney stone prediction is discussed with an exact tabulation to provide effectiveness.

   Keywords DL model, Kidney stone prediction, performance metrics, Accuracy, classification

   1. INTRODUCTION

      In the recent era, kidney stones are increased all over the world drastically which caused kidney failure and pain for humans. The kidney is the most essential organ that caused stone formation due to unwanted substances in urine [1-3]. Kidney stone diseases are now most common for all gender and sector peoples that too mostly in developed countries. The main reason for stone formation in kidneys is due to overweight, minimum level of water consumption, bad food diet and also regular intake of medicines and sometimes stress etc[4,5].

      The prediction of Kidney stone disease is mandatory due to its harmfulness. There are imaging techniques that are used for the diagnosis of a kidney stone to increase the lifespan of the affected person [6-9]. Some of the systems used to diagnose kidney stones are blood tests, urine tests, CT scans, ultrasonography, X-ray and MRI scans etc. These imaging techniques are easy to predict the kidney stone and simple the

      Doctor work with automation. This can be beneficial by saving time and minimising the risk of an error [10].

      To avoid false and error in prediction, Machine Learning (ML) and Deep Learning (DL) are implemented in medical imaging [11-15]. In the recent eras, ML and DL-based techniques are applied for so many fields such as Agriculture, communication, the Military, share market, education, weather forecasting and also in Medical Fields etc [16,17]. Several ML/DL methods are frequently used in medical applications for various issues such as Breast cancer, Brain tumour, Alzheimer's disease, Heart fat, Bone crack, Throat cancer, skin cancer, kidney stone and so on [18-20].

      Some of the popular ML/DL methods that are often used for medical imaging are Naive Bayes (NB), Decision Tree (DT), Neural Network (NN), Support Vector Machine (SVM), Random Forest (RF), Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM), AlexNet, GoogleNet, DenseNet, ShuffleNet, MobileNet and so on. These methods are very efficient in the prediction and also provided superior performance in segmentation and classification respectively [21-25].

      The rest of the work is structured as: section 2 presented the kidney stone symptoms and the types of stones are discussed in section 3. Section 4 presented the related work with tabulation and section 5 described the performance metrics. Section 6 presented the conclusion with a future enhancement in it.

   2. SYMPTOMS OF KIDNEY STONE

      The kidney stones are predicted by several symptoms that are listed below [26-28]:

      1. Renal colic that provided excessive cramping pain,

      2. Flank pain that caused in the backside,

      3. Hematuria which means blood in urine,

      4. Infection in the urinary tract,

      5. Obstructive uropathy means the disease occurred in the urine tract,

      6. Urine blockages,

      7. Hydronephrosis that was a kidney dilation

      Therefore, these affected the health and work of patients which required immediate and efficient treatment to enhance their quality of life [29,30].

   3. TYPES OF KIDNEY STONE

      Kidney stones are mainly caused due to a chemical abnormality in urine. There are several stones are formed in the human kidney based on various chemical compositions, sizes and shapes [31,32]. The most common type of stones is given in the following [33].

      1. Cystine Stones

         This type of stone is caused by a genetic disorder based on cystine and amino acid. It has occurred a 2% of kidney stone patients that provided a urinary excretion of excessive cystinuria. The cystine does not dissolve in urine and caused a cystine stone that evacuates 600 millimoles of cystine/per day[38].

      2. Drug-Induced Stones

      Some the drugs like triamterene, guaifenesin, atazanavir, and sulfa drugs are the reason for Drug-Induced stones [39]. It has occurred in 1% of kidney stone patients. For example, a frequent drug with a heavy dose for the disease especially for HIV would develop Drug-Induced Stones and leads to kidney failure [40].

      KIDNEY STONE

      A

      B

      C

      D

      E

      Fig 1. Classification of Kidney Stone

      1. Calcium Stones

         Calcium stones are the most common renal stones that are caused for many patients due to a urinary calculi abnormality. It is a mixture of Calcium Oxalate and Calcium Phosphate which is occurred in more than 60% of kidney stone patients. It has two types namely CaOx monohydrate (COM) and CaOx dihydrate (COD) where COM is frequently observed disease than COD [34].

      2. Struvite Stones

      Struvite stones are referred to as an infection that is caused by upto 1015% of patients. It is an ingredient of Magnesium Ammonium Phosphatewhich seems to be a triple phosphate stone. These stones caused various diseases like Klebsiella pneumonia, Enterobacter and Pseudomonas aeruginosa [35]. These stones have mostly occurred in women than males.

      C.Uric Acid Stones or Urate

      It has occurred in 310% of patients with kidney stones which comprised a high purine of animal protein like fish and meat. This stone has a possibility to cause a low urine volume, hyperuricosuria and low urinary pH. These types of uric acid stones are commonly occurred in males than in women [36,37].

   4. RELATED WORKS

      Several methods are developed by various authors in recent decades for kidney stone prediction that is discussed in the following and tabulated in Table 1.

      Black et al [41] presented the CNN and ResNet-101 model for a multi-class kidney stone classification. It can be diagnosed through automatic imaging based on cross- validation. This system has achieved a weighted recall of 94.12%, specificity of 97.83% and precision of 94.12% with an efficient performance. Besides, another automated stone detection is implemented by Yildirim et al [42] using 1799 images of coronal CT. The stone prediction is attained accurately even for a small size with an evaluation of 96.82% accuracy, 98% precision, 96% recall and 97% F1 score respectively.

      In some cases, the DL method of VGGNet-19 and Binary SVM is presented by Somasundaram et al [43] for effective feature extraction and classification of kidney stones. This system isperformed and validated with a97.32% Recall, 98.07% precision, 99.19 % accuracy, 98.28% specificity and93.2% F1 score respectively. Alongside, the ML/DL method predicted Renal Cell Carcinoma (RCC), urolithiasis, bladder cancer (BCa), and prostate cancer (PCa)based on the PubMed MEDLINE database by Suarez-Ibarrola et al [44]. It provided urological applications with a performance of 94.26% accuracy, 94.51% sensitivity and 96.29% specificity. Also, Artificial intelligence (AI) has been explored for time management to predict the tiny stones. This system can be diagnosed with a minimum CT image computation with a superior 98.52% accuracy, 96.75% sensitivity and 98.12% specificity by Sabuncu et al [45].

      Flores-Araiza et al [46] developed the part-prototypes (PPs) methodology with three categories namely the AlexNet model, batch normalization as the backbone (PPN-VGG19bn) and batch normalization of VGG19 (VGG19bn) respectively. The kidney stone classification is processed as a patch image and provided a mathematical analysis with an effective method. The performance of AlexNet attained a 96% accuracy, VGG19bn acquired 99% accuracy and also 98% accuracy for PPN-VGG19bn. Moreover, Li, D et al [47] investigated various DL methodssuch as 3D U-Net, SegNet, Res U-Net, DeepLabV3+, and UNETR for testing and training. These 3D U-Net, SegNet, Res U-Net, DeepLabV3+, and UNETR model has evaluated the accuracy of 98.32%, 97.45%, 89.93%,

      95.54% and 98.11% respectively.

      In some methods, the kidney stone predictions using an Inception v3 method by Ochoa-Ruiz et al [48] are tested with urinary calculi types in-vivo images. The performance metrics are evaluated with a 97% of weighted precision, 98% of recall and 97% of F1-score correspondingly. Moreover, the Renal Stone Complexity of the Seoul National University (S-ReSC) model is implemented by Jeong et al [49] for stone-free rate prediction of a kidney. The result of the Area Under Curve (AUC) is achieved 0.860 for an effective prediction.

      Kim et al [50] aimed urolithiasis prediction based on seven DL models with detailed chemical composition. This method has acquired 1,332 stone images and executed four classes for every DL model. Overall, the Xception DL model performed superior toall others evaluated for all four classes of precision and recall as class 1: 94.24%, 91.73%, class 2:

      85.42%, 96.14%, class 3: 86.86%, 99.59% and class 4: 94.96%,

      98.82%.Also in another work, the kidney stone prediction survey of DL models such as SVM, NB, and NN are investigated by Sri et al [51]. This survey executed the comparison of performance accuracy with every model's effectiveness.

      Author	Methods	Metrics evaluation	Strength	Weakness
      Black et	CNN and	Recall-	Multi	flexibility
      al [41]	ResNet-	94.12%	class	and speedy
      	101	Specificity-	based	diagnosis
      	model for	97.83%,	Performa
      	a multi-	Precision-	nce
      	class	94.12%
      	kidney
      	stone
      	classificat
      	ion

      The DL method like multi-view AlexNet max and multi-View VGGnet max is used to extract the fuse of the various viewpoints for discriminant object features. This system presented by Villalvazo-Avila et al [52] presents a deep-learning method for extracting and fusing image information acquired from different viewpoints for discriminant object features. The result is validated for precision and Recall for multi-view AlexNet max are 95% and 94% and also for multi-View VGG16 max are 94% and 94% respectively. Besides, Kazemi et al [53] developed an ensemble learning for a robust kidney stone prediction with a 97.1% accuracy. This system metrics are acquired with an effective classification result such as precision of 97.1%, recall of 97.1%, F1 score of 97.1% and 99.6% AUC.

      Shah et al [54] reviewed AI and its application for the classification of cystoscopic, renal masses diagnosis using an MRI. This work established the reaction of treatment, prognosis, survival and recurrence of genomic and biomarker studies. Alongside, a few DL methods like ensemble model, Logistic Regression (LR) and RF models are used to predict kidney stones using CT images. This system is investigated by Kolli et al [55]. This can be addressed by the metrics in terms of 97.12% precision, 96.83% accuracy and 98.1% Recall respectively. In a few literatures, the abdominal CT dataset is used to segment and classify for both testing and training of

      kidney stones prediction. The AI driven diagnostic approaches

      are explored by Li, D et al [56] with a performance accuracy of

      95%, a sensitivity of 88% and a specificity of 91% respectively.

      The CNN model-based automatic kidney stone prediction using CT images is implemented by GP et al [57]. This system achieved an effective performance with an accuracy of 96.82%, Recall of 93.22%which is competent enough as distinguished it from previous algorithms. Moreover, Surya et al [58] developed a Backpropagation network and Fuzzy Clustering Mean Algorithm for various steps for medical assistance in kidney stone prediction. These methods are provided with animage pixel with a higher stone prediction accuracy of 97.92%, specificity of 95.79% and sensitivity of 97.23% correspondingly.

      The Renal stone diagnosis is presented in radiography based on renal stone disease developed by McCarthy et al [59]. This system used a STONE PLUS prediction tool with innovative management and diagnosis. It provides the best care for patients with a precision of 92%, Recall of 93.41% and sensitivity of 94.21%. Also, Buvaneswari et al [60] discussed a hybrid ButterflyNet and InceptionNet model for kidney stone identification. The performance metrics acquired a 90% F1 Score, 84% accuracy, 88% Recall and 94% precision which provided a superior result than the conventional.

      Yildirim	CT based	Accuracy-	Accurate	Connectivi
      et al [42]	automate	96.82%,	detection	ty
      	d stone	precision-	even for
      	detection	98%,	small
      		Recall-96%,	size
      		F1 score-	stone
      		97%
      Somasun	VGGNet-	Recall-	QoS	Energy
      daram et	19 for	97.32%,		consumpti
      al [43]	feature	precision-		on
      	extraction	98.07%,
      	and	accuracy-
      	Binary	99.19
      	SVM for	%,specificit
      	classificat	y-98.28%
      	ion	F1 score-
      		93.2%
      Suarez-	PubMed	Accuracy-	Multi-	Complexit
      Ibarrola	MEDLIN	94.26%,	subject	y
      et al [44]	E	sensitivity-	<>detection
      	database-	94.51%,
      	based	specificity-
      	RCC,	96.29%
      	BCa, PCa
      	predictio
      	n
      Sabuncu	AI-based	Accuracy-	Computa	Less
      et al [45]	CT	98.52%,	tional	Availabilit
      	images	sensitivity-	accuracy	y
      		96.75%,
      		specificity-
      		98.12%
      Flores-	PPN-	All three	Effective	Display
      Araiza et	VGG19b	methods	mathema	quality
      al [46]	n,	attained	tical
      	AlexNet,	Accuracy of	analysis
      	VGG19b	96%, 99%
      	n	and 98%
      Li, D et	3D U-	Accuracy-	Classific	Energy
      al [47]	Net,	98.32%,	ation	consumpti
      	SegNet,	97.45%,89.9	accuracy	on
      	Res U-	3%, 95.54%
      	Net,	and 98.11%
      	DeepLab
      	V3+, and
      	UNETR
      Ochoa-	Inception	Precision-	Low cost	Low
      Ruiz et	v3	97%,	and	Flexibility
      al [48]	method	Recall-98%,	simple	and speed
      		F1-score-
      		97%
      Jeong et al [49]	DL based S-ReSC	AUC-0.860	Performa nce	Avoid false

      	predictio n			diagnosis
      Kim et al [50]	7model used for kidney stone predict	Precision- 94.96%, Recall- 98.82%	QoS	Limited functionali ties
      Sri et al [51]	Survey of DLmetho d based kidney stone predictio n	–	comparis on of performa nce accuracy	Future ideas
      Villalvaz o-Avila et al [52]	Multi- view AlexNet max and multi- View VGGnet max	AlexNet max Recall- 95%, VGGnet max Recall- 94%	Complex ity	Energy consumpti on
      Kazemi et al [53]	ensemble learning model	Recall- 97.1%, F1 score- 97.1%, AUC-99.6%	Lifetime	Reliability, Efficiency
      Shah et al [54]	AI model survey	–	Compari son of diagnosis ,methods and materials	Numerical compariso n
      Kolli et al [55]	Ensemble model, LR and RF models for predictio n	Precision- 97.12%, Accuracy- 96.83%, Recall- 98.1%	Memory	Utility, Accuracy
      Li, D et al [56]	AI drive n diagnosti c approach es	Accuracy- 95%, Sensitivity- 88%, Specificity 91%	Performa nce	Training
      GP et al [57]	CNN model	Accuracy- 96.82%, Recall- 93.22%	Cost overhead	Microscopi c examinatio n

      Surya et	Back	Accuracy-	Compres	Quite
      al [58]	propagati	97.92%,	sion and	complex
      	on	Specificity-	low data	system
      	network	95.79%,	loss
      	and	Sensitivity-
      	Fuzzy	97.23%
      	Clusterin
      	g Mean
      McCarth	Renal	Precision-	Effective	Security
      y et al	stone	92%,	ness
      [59]	diagnosis	Recall-
      		93.41%,
      		Sensitivity-
      		94.21%
      Buvanes	hybrid	F1 Score-	Portabilit	Categorise
      wari et	Butterfly	90%,	y, real-	the data
      al [60]	Net and	Accuracy-	time	Layer
      	Inception	84%,	Monitori
      	Net	Recall-88%,	ng
      	model	Precision-
      		94%

      (4)

      Recall: It is only concentrated only on Falsely Negative and won't be considered Truly Negatives which is expressed as equation (5).

      (5)

      Precision: It is measured only the Falsely Positives and Truly Positives that are expressed as equation (6).

   5. PERFORMANCE METRICS

      In the medical field, the performance metrics based on classifications are validated in terms of Sensitivity, Accuracy, Specificity, AUC and F1 score respectively. All these metrics are discussed and expressed below.

      Sensitivity: It is defined as actual positive cases proportional measurement which is truly detected as positive that is expressed in equation (1).

      (1)

      Specificity: It is the measure of actual negative proportions that is truly predicted as negative which is expressed in equation (2).

      (2)

      Accuracy: It is defined as the number of truly predicted divided by the overall predictions that are expressed in equation (3).

      (3)

      F1 Score: It measured the accuracy of testing which evaluated the precision and recall values that havethe best value as 1 and the worst value as 0which is expressed as equation (4).

      (6)

      Where indicates actually predicted as 1, denotes actually predicted as 0, represents the original value as 0 and detected as

      1 and indicates the original value as 1 and detected as 0 respectively.

   6. CONCLUSION

      In this article, medical image processing is reviewed for kidney stone prediction. A few kidney stone prediction works are listed and tabulated based on their method, achievement, merits and demerits. From this work, there are detailed explanations are given for every paper with its functions. It is clear to update the recent innovations towards kidney stone prediction. Though there are several challenges that are also still presented in prediction one has to build the QoS of a systemby an effective prediction by implementing some optimization algorithm with DL methods for fine-tuned results. Also, the security is not yet improved in these works, so the security has to be concentrated in future enhancements.

   7. REFERENCES

[1] Alina Jade Barnett, Fides Regina Schwartz, Chaofan Tao, Chaofan Chen, Yinhao Ren, Joseph Y. Lo, and Cynthia Rudin. IAIA-BL: A case-based interpretable deep learning model for classification of mass lesions in digital mammography. CoRR, abs/2103.12308, 2021.

[2] Kristian M Black, Hei Law, Ali Aldoukhi, Jia Deng, and Khurshid R Ghani. Deep learning computer vision algorithm for detecting kidney stone composition. 2020.

[3] Chaofan Chen, Oscar Li, Chaofan Tao, Alina Jade Barnett, Jonathan Su, and Cynthia Rudin. This Looks Like That: Deep Learning for Interpretable Image Recognition. CoRR, jun 2018.

[4] Mariela Corrales, Steeve Doizi, YazeedBarghouthy, Olivier Traxer, and Michel Daudon. Classification of stones according to micheldaudon: a narrative review. European Urology Focus, 7(1):1321, 2021.

[5] Michel Daudon, Arnaud Dessombz, Vincent Frochot, Emmanuel Letavernier, Jean-Philippe Haymann, Paul Jungers, and Dominique Bazin. Comprehensive morphoconstitutional analysis of urinary stones improves etiological diagnosis and therapeutic strategy of nephrolithiasis. ComptesRendusChimie, 19(11-12):14701491, 2016.

[6] Michel Daudon, Paul Jungers, Dominique Bazin, and James C Williams. Recurrence rates of urinary calculi according to stone composition and morphology. Urolithiasis, 46(5):459 470, 2018.

[7] Jonathan El Beze, Charles Mazeaud, Christian Daul, Gilberto Ochoa-Ruiz, Michel Daudon, Pascal Eschwege, ` and Jacques Hubert. Evaluation and understanding of automated urolithiasis recognition methods. BJU International, 2022.

[8] Laurence Estepa and Michel Daudon. Contribution of fourier transform infrared spectroscopy to the identification of urinary stones and kidney crystal deposits. Biospectroscopy, 3(5):347369, 1997.

[9] Vincent Estrade, Michel Daudon, Emmanuel Richard, JeanChristophe Bernhard, Franck Bladou, Gregoire Robert, and Baudouin Denis de Senneville. Towards automatic recognition of pure & mixed stones using intraoperative endoscopic digital images. BJU International, abs/2105.10686, 2021.

[10] Justin I Friedlander, Jodi A Antonelli, and Margaret S Pearle. Diet: from food to stone. World journal of urology, 33(2):179185, 2015.

[11] Francisco Lopez, Andres Varelo, Oscar Hinojosa, Mauricio Mendez, Dinh-Hoan Trinh, YonathanElBeze, Jacques Hubert, Vincent Estrade, Miguel Gonzalez, Gilberto Ochoa, et al. Assessing deep learning methods for the identification of kidney stones in endoscopic images. In 2021 43rd Annual International Conference of the IEEE Engineering in Medicine

& Biology Society (EMBC), pages 27782781. IEEE, 2021.

[12] Adriana Mart´nez, Dinh-Hoan Trinh, Jonathan El Beze, Jacques Hubert, Pascal Eschwege, Vincent Estrade, Lina Aguilar, Christian Daul, and Gilberto Ochoa. Towards an automated classification method for ureteroscopic kidney stone

images using ensemble learning. In 2020 42nd Annual International Conference of the IEEE Engineering in Medicine

& Biology Society (EMBC), pages 19361939. IEEE, 2020.

[13] Leland McInnes, John Healy, and James Melville. Umap: Uniform manifold approximation and projection for dimension reduction. arXiv preprint arXiv:1802.03426, 2018.

[14] Anmar Nassir, Hesham Saada, Taghreed Alnajjar, Jomanah Nasser, Waed Jameel, SohaElmorsy, and HattanBadr. The impact of stone composition on renal function. Urology Annals, 10(2):215, 2018.

[15] MeikeNauta, Annemarie Jutte, Jesper C. Provoost, and Christin Seifert. This looks like that, because … explaining prototypes for interpretable image recognition. CoRR, abs/2011.02863, 2020.

[16] Gilberto Ochoa-Ruiz, Vincent Estrade, Francisco Lopez, Daniel Flores-Araiza, Jonathan El Beze, Dinh-Hoan Trinh, Miguel Gonzalez-Mendoza, Pascal Eschwege, Jacques Hu- ` bert, and Christian Daul. On the in vivo recognition of kidney stones using machine learning. arXiv preprint arXiv:2201.08865, 2022.

[17] DawidRymarczyk, Lukasz Struski, Jacek Tabor, and Bartosz Zielinski. Protopshare: Prototype sharing for interpretable image classification and similarity discovery. CoRR, abs/2011.14340, 2020.

[18] Joan Serrat, Felipe Lumbreras, Francisco Blanco, Manuel Valiente, and Montserrat Lopez-Mesas. mystone: A sys- ´ tem for automatic kidney stone classification. Expert Systems with Applications, 89:4151, 2017.

[19] R Siener and A Hesse. Fluid intake and epidemiology of urolithiasis. European journal of clinical nutrition, 57(2):S47 S51, 2003.

[20] Alejandro Torrell Amado. Metric learning for kidney stone classification. 2018. 2 [21] Adie Viljoen, Rabia Chaudhry, and John Bycroft. Renal stones. Annals of clinical biochemistry, 56(1):1527, 2019.

[21] Lemaitre, G., Nogueira, F., Aridas, C., 2017. Imbalanced- learn: a python toolbox to tackle the curse of imbalanced datasets in machine learning. Journal of Machine Learning Research 18, 15.

[22] Martínez, A., Trinh, D.H., El Beze, J., Hubert, J., Eschwege, P., Estrade, V., Aguilar, L., Daul, C., Ochoa, G., 2020. Towards an automated classification method for

ureteroscopic kidney stone images using ensemble learning, in: 2020 42nd Annual International Conference of the IEEE Engineering in Medicine Biology Society (EMBC), pp. 1936 1939.

[23] McCarthy, C.J., Baliyan, V., Kordbacheh, H., Sajjad, Z., Sahani, D., Kambadakone, A., 2016. Radiology of renal stone disease. International Journal of Surgery 36, 638646.

[24] McInnes, L., Healy, J., Melville, J., 2020. Umap: Uniform manifold approximation and projection for dimension reduction. arXiv:1802.03426.

[25] Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Prettenhofer, P., Weiss, R., Dubourg, V., Vanderplas, J., Passos, A., Cournapeau, D., Brucher, M., Perrot, M., Duchesnay, E., 2011. Scikit-learn: Machine learning in Python. Journal of Machine Learning Research 12, 28252830.

[26] Sampogna, G., Basic, D., Geavlete, P., GalánLlopis, J., Reis Santos, J., Saltirov, I., Sarica, K., Stavridis, S., Skolarikos, A., Trinchieri, A., 2020. Identificaciónendoscópica de la composición de loscálculosurinarios: un estudio del southeastern group for lithiasis research (segur 2).

[27] Scales, C.D., Smith, A.C., Hanley, J.M., Saigal, C.S., 2012. Prevalence of kidney stones in the united states. European Urology 62, 160 165.

[28] Selvaraju, R.R., Cogswell, M., Das, A., Vedantam, R., Parikh, D., Batra, D., 2017. Grad-cam: Visual explanations from deep networks via gradientbased localization, in: 2017 IEEE International Conference on Computer Vision (ICCV), pp. 618626.

[29] Serrat, J., Lumbreras, F., Blanco, F., Valiente, M., López- Mesas, M., 2017. mystone: A system for automatic kidney stone classification. Expert Systems with Applications 89, 41 51.

[30] Siener, R., Buchholz, N., Daudon, M., Hess, B., Knoll, T., Osther, P.J., ReisSantos, J., Sarica, K., Traxer, O., Trinchieri, A., of Urolithiasis (EULIS), E.S., 2016. Quality assessment of urinary stone analysis: Results of a multicenter study of laboratories in europe. PLOS ONE 11, 110.

[31] D. P. Griffith, Struvite stones, Kidney International, vol. 13, no. 5, pp. 372382, 1978.

[32] M. Dursun, A. Otunctemur, and E. Ozbek, Kidney stones and ceftriaxone, European Medical Journal of Urology, vol. 3, no. 1, pp. 6874, 2015.

[33] F. L. Coe, A. Evan, and E. Worcester, Kidney stone disease, Journal of Clinical Investigation, vol. 115, no. 10, pp. 25982608, 2005.

[34] Chaudhary, S. K. Singla, and C. Tandon, In vitro evaluation of Terminalia arjuna on calcium phosphate and calcium oxalate crystallization, Indian Journal of Pharmaceutical Sciences, vol. 72, no. 3, pp. 340345, 2010.

[35] F. L. Coe, J. H. Parks, and J. R. Asplin, The pahogenesis and treatment of kidney stones, New England Journal of Medicine, vol. 327, no. 16, pp. 11411152, 1992.

[36] A. Skolarikos, M. Straub, T. Knoll et al., Metabolic evaluation and recurrence prevention for urinary stone patients: EAU guidelines, European Urology, vol. 67, no. 4, pp. 750 763, 2015.

[37] A. Bensatal and M. R. Ouahrani, Inhibition of crystallization of calcium oxalate by the extraction of Tamarix gallica L, Urological Research, vol. 36, no. 6, pp. 283287, 2008.

[38] D. R. Basavaraj, C. S. Biyani, A. J. Browning, and J. J. Cartledge, The role of urinary kidney stone inhibitors and promoters in the pathogenesis of calcium containing renal stones, EAU-EBU Update Series, vol. 5, no. 3, pp. 126136, 2007.

[39] F. Dal-Moro, M. Mancini, I. M. Tavolini, V. De Marco, and P. Bassi, Cellular and molecular gateways to urolithiasis: a new insight, Urologia Internationalis, vol. 74, pp. 193197, 2005.

[40] D. V. Kishore, F. Moosavi, and D. R. K. Varma, Effect of ethanolic extract of Portulaca oleracea linn. on ethylene glycol and ammonium chloride induced urolithiasis, International Journal of Pharmacy and Pharmaceutical Sciences, vol. 5, no. 2, pp. 134140, 2013.

[41] Black, K. M., Law, H., Aldoukhi, A., Deng, J., & Ghani,

K. R. (2020). Deep learning computer vision algorithm for detecting kidney stone composition. BJU international, 125(6), 920-924.

[42] Yildirim, K., Bozdag, P. G., Talo, M., Yildirim, O., Karabatak, M., & Acharya, U. R. (2021). Deep learning model for automated kidney stone detection using coronal CT images. Computers in biology and medicine, 135, 104569.

[43] Somasundaram, K., & Sivakumar, P. (2021, December). An Efficient Detection of Kidney Stone Based on HDVS Deep Learning Approach. In Proceedings of the First International Conference on Combinatorial and Optimization, ICCAP 2021, December 7-8 2021, Chennai, India.

[44] Suarez-Ibarrola, R., Hein, S., Reis, G., Gratzke, C.,

&Miernik, A. (2020). Current and future applications of machine and deep learning in urology: a review of the literature on urolithiasis, renal cell carcinoma, and bladder and prostate cancer. World journal of urology, 38(10), 2329-2347.

[45] Ö. Sabuncu and B. Bilgehan, "Performance Evaluation for Various Deep Learning (DL) Methods Applied to Kidney Stone Diseases," 2021 International Conference on Forthcoming Networks and Sustainability in AIoT Era (FoNeS- AIoT), 2021, pp. 1-3, doi: 10.1109/FoNeS- AIoT54873.2021.00010.

[46] Flores-Araiza, D., Lopez-Tiro, F., Villalvazo-Avila, E., El- Beze, J., Hubert, J., Ochoa-Ruiz, G., &Daul, C. (2022). Interpretable Deep Learning Classifier by Detection of Prototypical Parts on Kidney Stones Images. arXiv preprint arXiv:2206.00252.

[47] Li, D., Xiao, C., Liu, Y., Chen, Z., Hassan, H., Su, L., … & Huang, B. (2022). Deep Segmentation Networks for Segmenting Kidneys and Detecting Kidney Stones in Unenhanced Abdominal CT Images. Diagnostics, 12(8), 1788.

[48] Ochoa-Ruiz, G., Estrade, V., Lopez, F., Flores-Araiza, D., Beze, J. E., Trinh, D. H., … &Daul, C. (2022). On the in vivo recognition of kidney stones using machine learning. arXiv preprint arXiv:2201.08865.

[49] Jeong, C. W., Jung, J. W., Cha, W. H., Lee, B. K., Lee, S.,

Jeong, S. J., … & Lee, S. E. (2013). Seoul National University Renal Stone Complexity Score for predicting stone-free rate after percutaneous nephrolithotomy. PLoS One, 8(6), e65888.

[50] Kim, U. S., Kwon, H. S., Yang, W., Lee, W., Choi, C.,

Kim, J. K., … & Han, J. H. (2021). Prediction of the composition of urinary stones using deep learning. Investigative and Clinical Urology, 63.

[51] Sri, V. S., Kumar, P. S., & Rajendran, V. (2023). A Review on Detection of Kidney Disease Using Machine Learning and Deep Learning Techniques. Application of Deep Learning Methods in Healthcare and Medical Science, 1-22.

[52] Villalvazo-Avila, E., Lopez-Tiro, F., Flores-Araiza, D., Ochoa-Ruiz, G., El-Beze, J., Hubert, J., &Daul, C. (2022).

Comparing feature fusion strategies for Deep Learning-based kidney stone identification. arXiv preprint arXiv:2206.00069.

[53] Kazemi, Y., &Mirroshandel, S. A. (2018). A novel method for predicting kidney stone type using ensemble learning. Artificial intelligence in medicine, 84, 117-126.

[54] Shah, M., Naik, N., Somani, B. K., & Hameed, B. Z. (2020). Artificial intelligence (AI) in urology-Current use and future directions: An iTRUE study. Turkish Journal of Urology, 46(Suppl 1), S27.

[55] C. S. Kolli, M. P. Raghunath, S. Meenakshi, K. Maheswari, C. F. Britto and S. Kushwaha, "Efficient Development of Supervised Learning Algorithm for kidney Stone Prediction," 2022 International Conference on Inventive Computation Technologies (ICICT), 2022, pp. 1373-1379, doi: 10.1109/ICICT54344.2022.9850573.

[56] Li, D., Xiao, C., Liu, Y., Chen, Z., Hassan, H., Su, & Zhong, W. (2022). Deep Segmentation Networks for Segmenting Kidneys and Detecting Kidney Stones in Unenhanced Ab-dominal CT Images. Diagnostics 2022, 12, 1788.

[57] GP, V. P., Reddy, K. V. S., Kiruthik, A. M., &ArunNehru,

J. Prediction of Kidney Stones Using Machine Learning.

[58] Surya v, P.V.Sumanth , U.Surendra , V.Chakradhar, (2021). Kidney Stone Detection using Machine Learning Algorithm, Vol XIV Issue 5 2021, ISSN NO : 0090-5674.

[59] McCarthy, C. J., Baliyan, V., Kordbacheh, H., Sajjad, Z., Sahani, D., &Kambadakone, A. (2016). Radiology of renal stone disease. International Journal of Surgery, 36, 638-646.

[60] Buvaneswari, R., & Vinoth, R. (2022). Kidney Stone Detection Using Hybrid Butterfly Net and Inception net Model. ESP Journal of Engineering and Technology Advancements, 2(4), 1-6.