Musculoskeletal (MSK) disorders, such as fractures, degenerative diseases like osteoarthritis, and soft tissue injuries, are widespread health concerns that often require imaging for accurate diagnosis and treatment planning. Traditional imaging techniques, including X-rays, Computed Tomography (CT), and Magnetic Resonance Imaging (MRI), depend on the expertise of radiologists for image interpretation. However, this process can be time-consuming and is susceptible to variability and human error. In recent years, deep learning, a subset of Artificial Intelligence (AI), has emerged as a promising tool for automating and enhancing the interpretation of medical images. Specifically, Convolutional Neural Networks (CNNs) have shown remarkable success in identifying and classifying various MSK conditions with high precision. Deep learning applied to MSK imaging: A narrative review examining current approaches in the key initiatives of image classification, segmentation, and abnormality detection CNNs have shown great promise in accurate fracture diagnosis, earlier identification of osteoarthritis, and the distinction between soft tissue injuries.