Abstract—Kidney stone detection is a critical task in medical imaging due to its significant impact on patient treatment. Early detection of kidney stones is crucial, as untreated cases can cause complications. However, machine learning methods for detecting kidney stones in medical imaging face challenges such as data availability, model complexity, interpretability, generalization, and ethical considerations. This study introduces an innovative hybrid approach that combines deep learning with optimization techniques to enhance kidney stone detection. The proposed method integrates watershed segmentation to improve feature extraction and facilitate more precise delineation of object boundaries, enhancing the quality of extracted features. This method uses a Convolutional Neural Network (CNN) to extract features and a Deep Belief Network (DBN) for classification. The CNN learns spatial hierarchies and edges, while the DBN has a multilayer structure for effective classification. Additionally, a hybrid optimization scheme that integrates Crow with Cuckoo Search Optimizer is introduced to further enhance the model’s performance. This approach outperforms existing methods such as Resnet-50, DA-CNN, and Deep Learning (DL) with ResNet, achieving a remarkable accuracy of 99.35% in kidney stone detection. It significantly increases the accuracy of kidney stone detection due to the creative combination of data pre-processing, feature extraction, and hybrid Marker-based segmentation with a Crow-Cuckoo Search Optimizer (CCSO). Detecting kidney stones using existing methods such as Magnetic Resonance Imaging (MRI), Ultrasounds, and X-ray imaging faces challenges such as extended scanning times, invasiveness, radiation exposure, and inaccuracies. This study emphasizes using deep learning and optimization techniques to improve kidney stone detection and diagnostic accuracy in medical imaging, leading to better patient care.
 
Keywords—kidney stone detection, watershed segmentation, hybrid crow-cuckoo search optimizer, convolutional neural network, deep belief network


Copyright © 2025 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (CC BY 4.0).