Evaluation of Customized ACL-Net Model: A Comparative Analysis of CNN and Image Enhancement Techniques for Classifying ACL Injuries

Magnetic resonance imaging (MRI) plays a crucial role in diagnosing knee injuries and joint conditions, particularly in detecting anterior cruciate ligament (ACL) damage that can impact knee stability. This research employed various convolutional neural network (CNN) models, including the customized ACL-Net, to classify ACL conditions in MRI image datasets, applying image processing techniques for dataset enhancement. ACL-Net outperformed, achieving 93.75%, 91.56%, 87.44%, and 90.65% accuracy for full-sized, enhanced, ROI-cropped, and ROI-cropped enhanced MRI images, respectively. The findings suggest that ACL-Net is a promising tool for automated knee ligament injury classification in MRI images, with clinical diagnostic potential.