Hybrid CNN-Attention-LSTM Architecture for Automated Kidney Stone Detection and Classification from Medical Imaging
Downloads
There are many common diseases that affect millions of people around the world, and among these diseases is kidney stones (urinary tract stones). Its detection requires accurate and early diagnosis in order for appropriate treatment to be carried out as quickly as possible, because delaying it causes the patient’s condition to deteriorate. Traditional diagnostic methods also rely heavily on the radiologist, which is time-consuming. While this paper presents a new deep learning hybrid architecture that combines convolutional neural networks (CNNs), attention mechanisms, and long-term short-term memory networks (LSTM) in order to automatically detect and also classify kidney stones based on medical images. The proposed model also uses CNN to extract spatial features, attention mechanisms to focus on subject- related domains, and LSTM to capture sequential dependencies in multi-slice CT scans. The model was evaluated on a comprehensive and accurate dataset of 4850 CT images, an accuracy of 96.2%, an accuracy of 95.8%, a recall of 96.5%, and an F1 score of 96.1% were obtained in classification tasks. The attention mechanism performs the precise function of improving interpretability by highlighting areas of interest, making the system clinically applicable. Comparative analysis using state-of-the-art methods including ResNet50, VGG16, and traditional CNN architecture shows superior performance. The proposed architectural design also has an average AUC of 0.95 when applied to four types of stones (calcium oxalate, uric acid, struvite, and cysteine), showing a significant superiority over existing methods of 3-5.
A. Parakh, H. Lee, J. H. Lee, B. H. Eisner, D. V. Sahani, and S. Do, “Urinary stone detection on CT images using deep convolutional neural networks: evaluation of model performance and generalization,” Radiology: Artificial Intelligence, vol. 1, no. 4, p. e180066, 2019.
D. C. Elton, E. B. Turkbey, P. J. Pickhardt, and R. M. Summers, “A deep learn- ing system for automated kidney stone detection and volumetric segmentation on noncontrast CT scans,” Medical Physics, vol. 49, no. 4, pp. 2545–2554, 2022.
M. Kobayashi, J. Ishioka, Y. Matsuoka, Y. Fukuda, Y. Kubota, K. Saito, Y. Fu- jii, and K. Kihara, “Computer-aided diagnosis with a convolutional neural network algorithm for automated detection of urinary tract stones on plain X-ray,” BMC Urology, vol. 21, no. 1, pp. 1–9, 2021.
T. De Perrot, J. Hofmeister, S. Burgermeister, A. Poletti, X. Montet, and S. Becker, “Differentiating kidney stones from phleboliths in unenhanced low-dose computed tomography using radiomics and machine learning,” European Radiology, vol. 29, no. 9, pp. 4776–4782, 2019.
H. E. Kim, A. Cosa-Linan, N. Santhanam, M. Jannesari, M. E. Maros, and T. Ganslandt, “Transfer learning for medical image classification: a literature review,” BMC Medical Imaging, vol. 22, no. 1, pp. 1–13, 2022.
O. N. Manzari, H. Ahmadabadi, H. Kashiani, S. B. Shokouhi, and A. Ayatollahi, “MedViT: a robust vision transformer for generalized medical image classification,” Computers in Biology and Medicine, vol. 157, p. 106791, 2023.
A. Caglayan, M. O. Horsanali, K. Kocadurdu, E. Ucar, and I. Guney, “Deep learning model-assisted detection of kidney stones on computed tomography,” International Brazilian Journal of Urology, vol. 48, no. 6, pp. 1067–1075, 2022.
F. S. Tahir and A. A. Abdulrahman, “Kidney stones detection based on deep learning and discrete wavelet transform,” Indonesian Journal of Electrical Engineering and Computer Science, vol. 30, no. 2, pp. 1104–1112, 2023.
I. Aksakallı, S. Ka¸cdıog˘lu, and Y. S. Hanay, “Kidney x-ray images classification using machine learning and deep learning methods,” Balkan Journal of Electrical and Computer Engineering, vol. 9, no. 2, pp. 144–151, 2021.
T. Gon¸calves, I. Rio-Torto, L. F. Teixeira, and J. S. Cardoso, “A survey on attention mechanisms for medical applications: are we moving toward better algorithms?” IEEE Access, vol. 10, pp. 98909–98935, 2022.
I. Salehin, M. S. Islam, N. Amin, M. A. Baten, M. S. Hossain, and A. Matin, “Real- time medical image classification with ML framework and dedicated CNN–LSTM architecture,” Journal of Healthcare Engineering, vol. 2023, Article ID 3717035, 2023.
M. M. Srikantamurthy, V. P. S. Rallabandi, D. B. Dudekula, N. Natarajan, and S. Park, “Classification of benign and malignant subtypes of breast cancer histopathol- ogy imaging using hybrid CNN-LSTM based transfer learning,” BMC Medical Imag- ing, vol. 23, no. 1, pp. 1–17, 2023.
J. Jendeberg, P. Thunberg, and M. Lid´en, “Differentiation of distal ureteral stones and pelvic phleboliths using a convolutional neural network,” Urolithiasis, vol. 49, no. 1, pp. 41–47, 2021.
F. Lopez-Tiro, V. Estrade, J. Hubert, J. Traxer, M. Daudon, H. Shu, D. Sarocchi, and C. Daul, “On the in vivo recognition of kidney stones using machine learning,” IEEE Access, vol. 12, pp. 10211–10232, 2024.
S. J. Eun, M. S. Yun, T. K. Whangbo, Y. S. Moon, J. M. Choi, and S. E. Lee, “A study on the optimal artificial intelligence model for determination of urolithiasis,” International Neurourology Journal, vol. 26, no. 3, pp. 202–210, 2022.
