Performance Analysis of Deep Learning Models in the Diagnosis of Kidney Diseases

Öz

Chronic kidney disease is a major global health problem affecting 10-15% of the world's population and can lead to serious health problems if left untreated. If not diagnosed early, it can lead to kidney failure and cardiovascular complications. In this study, the effectiveness of DL models in diagnosing and classifying CKD was evaluated using the CT-KIDNEY dataset. InceptionV3, ResNet50, CNN, CNN+LSTM and VGG16 architectures were compared using accuracy, precision, recall and F1 score metrics. The CNN+LSTM model showed the highest classification success with an accuracy of 91.0%. Transfer learning based models such as VGG16 (90.6%), InceptionV3 (89.6%) and ResNet50 (88.9%) also achieved competitive and high accuracy rates in the range of 88%-91%. In contrast, the standard CNN model was inadequate for this particular task, with a very low accuracy of 27.4%. The learning curves showed that the transfer learning models and CNN+LSTM exhibited successful learning processes, but there were tendencies to overlearn, especially in CNN and CNN+LSTM (CNN-LSTM also showed a high variance in test accuracy). ResNet50 showed the most balanced learning profile. Future work could aim at using larger and more diverse datasets, integrating them into IoT-based real-time diagnostic systems, and investigating their applicability in different medical imaging modalities.

Anahtar Kelimeler

• Chronic kidney disease
• Deep learning
• CT-KIDNEY
• Classification
• Convolutional Neural Networks
• Transfer Learning

Böbrek Hastalıklarının Tanısında Derin Öğrenme Modellerinin Performans Analizi

Öz

Kronik böbrek hastalığı, dünya nüfusunun %10-15’ini etkileyen ve tedavi edilmediğinde ciddi sağlık sorunlarına yol açabilen önemli bir küresel sağlık problemidir. Erken teşhis edilmediğinde böbrek yetmezliği ve kardiyovasküler komplikasyonlara neden olabilir. Bu çalışmada, KBH teşhis ve sınıflandırmasında DL modellerinin etkinliği, CT-KIDNEY veri seti kullanılarak değerlendirilmiştir. InceptionV3, ResNet50, CNN, CNN+LSTM ve VGG16 mimarileri doğruluk, kesinlik, hatırlama ve F1-skor metrikleri ile karşılaştırılmıştır. CNN+LSTM modelinin %91,0 doğruluk ile en yüksek sınıflandırma başarısını gösterdiğini ortaya koymuştur. VGG16 (%90,6), InceptionV3 (%89,6) ve ResNet50 (%88,9) gibi transfer öğrenme tabanlı modeller de %88-%91 aralığında rekabetçi ve yüksek doğruluk oranları elde etmiştir. Buna karşın, standart CNN modeli %27,4 gibi oldukça düşük bir doğruluk oranı sergileyerek, bu spesifik görev için yetersiz kalmıştır. Öğrenme eğrileri transfer öğrenme modellerinin ve CNN+LSTM'in başarılı öğrenme süreçleri sergilediğini, ancak özellikle CNN ve CNN+LSTM'de aşırı öğrenme eğilimleri (CNN-LSTM'de ayrıca test doğruluğunda yüksek varyans) olduğunu göstermiştir. ResNet50 en dengeli öğrenme profilini sunmuştur. Gelecekteki çalışmalarda, daha büyük ve çeşitli veri setleri kullanmayı, IoT tabanlı gerçek zamanlı tanı sistemlerine entegrasyonunu sağlamayı ve farklı tıbbi görüntüleme modalitelerinde uygulanabilirliğini incelemeyi hedefleyebilir.

Anahtar Kelimeler

• Kronik böbrek hastalığı
• Derin öğrenme
• CT-KIDNEY
• Sınıflandırma.
• Evrişimsel Sinir ağları
• Öğrenme Aktarımı

Kaynakça

1. Ahmed F, Abbas S, Athar A, Shahzad T, Khan WA, Alharbi M, ... Ahmed A (2024). Identification of kidney stones in KUB X-ray images using VGG16 empowered with explainable artificial intelligence. Scientific Reports, 14(1): 6173.
2. Akter R, Golam M, Doan VS, Lee JM, Kim DS (2022). Iomt-net: Blockchain-integrated unauthorized uav localization using lightweight convolution neural network for internet of military things. IEEE Internet of Things Journal, 10(8):6634-6651.
3. Ali L, Alnajjar F, Jassmi HA, Gocho M, Khan W, Serhani MA (2021). Performance evaluation of deep CNN-based crack detection and localization techniques for concrete structures. Sensors, 21(5):1688.
4. Alsuhibany SA, Abdel-Khalek S, Algarni A, Fayomi A, Gupta D, Kumar V, Mansour RF (2021). Ensemble of deep learning based clinical decision support system for chronic kidney disease diagnosis in medical internet of things environment. Computational Intelligence and Neuroscience, 2021(1):4931450.
5. Aydın ZBG, Şamlı R (2020). A comparison of software defect prediction metrics using data mining algorithms. Journal of Innovative Science and Engineering (JISE), 4(1):11-21.
6. Bingol H, Yıldırım M, Yıldırım K, Alatas B (2023). Automatic classification of kidney CT images with relief based novel hybrid deep model. PeerJ Computer Science, 9:e1717. doi:10.7717/peerj-cs.1717.
7. Debal DA, Sitote TM (2022). Chronic kidney disease prediction using machine learning techniques. Journal of Big Data, 9(1):109.
8. Doğru Ş, Altuntaş V (2023). Prediction of cancer in DNA sequences using unsupervised learning methods. Journal of Innovative Science and Engineering, 7(1):40-47.

9. Elhoseny M, Shankar K, Uthayakumar J (2019). Intelligent diagnostic prediction and classification system for chronic kidney disease. Scientific Reports, 9(1):9583.
10. Goel A (2022). Deployed deep learning kidney segmentation for polycystic kidney disease MRI. Radiol Artif Intell, 4(2):e210205.
11. He K, Zhang X, Ren S, Sun J (2016). Deep residual learning for image recognition. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 770-778.
12. Islam MN, Hasan M, Hossain MK, Alam MGR, Uddin MZ, Soylu A (2022). Vision transformer and explainable transfer learning models for auto detection of kidney cyst, stone and tumor from CT-radiography. Scientific Reports, 12(1):1-14.
13. Jeyalakshmi G, Lloyd FV, Subbulakshmi K, Vinudevi G (2024). Application of deep learning in identifying novel biomarkers for chronic kidney disease progression. In Proceedings of the 2024 10th International Conference on Advanced Computing and Communication Systems (ICACCS), vol. 1, 353–358. IEEE.
14. Khorram A, Khalooei M, Rezghi M (2021). End-to-end CNN+ LSTM deep learning approach for bearing fault diagnosis. Applied Intelligence, 51(2):736–751.
15. Kumar K, Pradeepa M, Mahdal M, Verma S, RajaRao MVLN, Ramesh JVN (2023). A deep learning approach for kidney disease recognition and prediction through image processing. Applied Sciences, 13(6): 3621.
16. Kumar S, Ratan R, Desai JV (2022). Cotton disease detection using tensorflow machine learning technique. Advances in Multimedia, 2022(1): 1812025.
17. Lal S, Chanchal AK, Kini J, Upadhyay GK (2024). FPGA implementation of deep learning architecture for kidney cancer detection from histopathological images. Multimedia Tools and Applications, 1-19.
18. Lee SH, Chan CS, Mayo SJ, Remagnino P (2017). How deep learning extracts and learns leaf features for plant classification. Pattern Recognition, 71: 1-13.
19. Ma F, Sun T, Liu L, Jing H (2020). Detection and diagnosis of chronic kidney disease using deep learning-based heterogeneous modified artificial neural network. Future Generation Computer Systems, 111: 17-26.
20. Mehedi MHK, Haque E, Radin SY, Rahman MAU, Reza MT, Alam MGR (2022). Kidney tumor segmentation and classification using deep neural network on CT images. 2022 International Conference on Innovations in Science, Engineering and Technology (ICISET), 406-411. doi: 10.1109/ICISET54810.2022.10034638.
21. Patro KK, Allam JP, Neelapu BC, Tadeusiewicz R, Acharya UR, Hammad M, ... Plawiak P (2023). Application of Kronecker convolutions in deep learning technique for automated detection of kidney stones with coronal CT images. Information Sciences, 640: 119005.
22. Rani S, Malu G, Sherly E (2023). Kidney stone detection from CT images using probabilistic neural network (PNN) and watershed algorithm. In 2023 International Conference on Advances in Intelligent Computing and Applications (AICAPS), 1-6. IEEE.
23. Sabanayagam C, Xu D, Ting DS, Nusinovici S, Banu R, Hamzah H, ... Wong TY (2020). A deep learning algorithm to detect chronic kidney disease from retinal photographs in community-based populations. The Lancet Digital Health, 2(6):e295-e302.
24. Saif D, Sarhan AM, Elshennawy NM (2023). Deep-kidney: an effective deep learning framework for chronic kidney disease prediction. Health Information Science and Systems, 12(1): 3.
25. Sapra P, Mary SSC, Chauhan A, Parte SA, Nishant N (2023). Hybrid convolutional neural network and extreme learning machine for kidney stone detection. In 2023 Second International Conference on Electronics and Renewable Systems (ICEARS), 936-942. IEEE.
26. Senthil K, Vidyaathulasiraman (2021). Ovarian cancer diagnosis using pretrained mask CNN-based segmentation with VGG-19 architecture. Bio-Algorithms and Med-Systems, (0): 000010151520210098.
27. Simonyan K, Zisserman A (2014). Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556.
28. Singh V, Asari VK, Rajasekaran R (2022). A deep neural network for early detection and prediction of chronic kidney disease. Diagnostics, 12(1):116.
29. Sulistyowati T, Purwanto P, Alzami F, Pramunendar RA (2023). VGG16 deep learning architecture using imbalance data methods for the detection of apple leaf diseases. Moneter: Jurnal Keuangan dan Perbankan, 11(1):41-53.
30. Sundaramoorthy S, Jayachandru K (2023). Designing of enhanced deep neural network model for analysis and identification of kidney stone, cyst, and tumour. SN Computer Science, 4(5):466.
31. Tahir FS, Abdulrahman AA (2023). Kidney stones detection based on deep learning and discrete wavelet transform. Indonesian Journal of Electrical Engineering and Computer Science, 31(3):1829.
32. Ul Hassan M (2018). Vgg16-convolutional network for classification and detection. Neurohive. Available at: https://neurohive.io/en/popular-networks/vgg16/ [Accessed 10.4.2019].
33. Venkatrao K, Kareemulla S (2023). HDLNET: a hybrid deep learning network model with intelligent IoT for detection and classification of chronic kidney disease. IEEE Access.
34. Wu J (2017). Introduction to convolutional neural networks. National Key Lab for Novel Software Technology, Nanjing University, China, 5(23):495.
35. Xia X, Xu C, Nan B (2017). Inception-v3 for flower classification. In: 2017 2nd International Conference on Image, Vision and Computing (ICIVC), 783-787, IEEE.
36. Yang J, Chen X, Luo C, Li Z, Chen C, Han S, ... Chen C (2023). Application of serum SERS technology combined with deep learning algorithm in the rapid diagnosis of immune diseases and chronic kidney disease. Scientific Reports, 13(1):15719.
37. Yogesh N, Shrinivasacharya P, Naik N (2024). Novel statistically equivalent signature-based hybrid feature selection and ensemble deep learning LSTM and GRU for chronic kidney disease classification. PeerJ Computer Science, 10:e2467.
38. Zhang L, Zhang J, Gao W, Bai F, Li N, Sheykhahmad FR (2024). A novel approach for automated diagnosis of kidney stones from CT images using optimized InceptionV4 based on combined dwarf mongoose optimizer. Biomedical Signal Processing and Control, 94:106356.
39. Zhang X, Agborbesong E, Li X (2021). The role of mitochondria in acute kidney injury and chronic kidney disease and its therapeutic potential. International Journal of Molecular Sciences, 22(20):11253.