A Novel CNN based Classification of Anterior Talofibular Ligament in T2 Axial Magnetic Resonance Imaging Scans

Plantar flexion and foot inversion contribute to about 70% of lateral ankle sprains, with anterior talofibular ligament (ATFL) experiencing the most in jury in these situations. For patients to treat and heal from ATFL injuries, a swift and accurate diagnosis is essential. However, radiologists’ traditional techniques of interpreting Magnetic Resonance Imaging (MRIs) can be labor-intensive and vulnerable to human error. To try to overcome these challenges, this research developed an advanced Convolutional Neural Network (CNN) model that automatically identifies ATFL MRI scans as either abnormal or normal. The Image processing method comprises an extensive setup process that includes augmentation, greyscale images transformation, applying Gaussian blur to minimize noise, and cropping using contour recognition. Images are scaled and normalized after preprocessing and then sent into CNN for training. When measured using many metrics, including precision, recall, and F1 score, the CNN model performed very well, achieving a diagnostic accuracy of 94.45%. This indicates that the CNN model is effective in distinguishing between normal and abnormal ATFL conditions. This prominent level of accuracy not only promises to significantly reduce the diagnostic workload for radiologists but also enhances the speed and reliability of the diagnostic process. The objective is to boost the application of this technology in clinical settings, which could result in better outcomes for patients, more effective diagnostic processes, improved medical imaging, and enhanced healthcare delivery in general. Future work will try to grow the dataset’s capacity to include a more diverse range of images and explore its potential application to the diagnosis of other ligament injuries.