Millions of people worldwide suffer from kidney stones, a common urological condition. It is essential to detect kidney stones early and accurately in order to plan effective treatments and minimize patient discomfort. To boost the accuracy of kidney stone detection, this study proposes a hybrid strategy that combines the classification capabilities of both Convolutional Neural Network (CNN) and the Support Vector Machine (SVM) classifier. CNN is used to extract features from images of kidney stones. After that, the features extracted are fed into an SVM classifier, which then decides whether kidney stones are present or not. SVM is chosen because it is effective in dealing with binary classification issues and can handle high-dimensional data. A dataset of kidney stone images with corresponding labels (stone presence or absence) is used to develop and evaluate the proposed method. The hybrid CNN-SVM approach's efficiency is evaluated by comparing it to the traditional machine learning approach. When compared to individual classifiers or other conventional methods, the suggested method outperforms them in terms of accuracy and classification performance. This study makes a contribution to the field of medical image analysis and provides a promising framework for the creation of automated kidney stone detection models that can help medical professionals better diagnose and treat patients.