This systematic review investigates the application of artificial intelligence (AI) and machine learning (ML) models in diagnosing and classifying hip fractures using radiographic imaging. Fourteen studies published up to June 2022 were analyzed, focusing on performance metrics like diagnostic accuracy, AUC (area under the ROC curve), and classification precision. The studies utilized X-rays, with some incorporating CT and MRI, examining fracture types such as femoral neck and intertrochanteric fractures. AI diagnostic accuracy ranged from 79.3% to 98%, with AUC values between 0.905 and 0.99, demonstrating excellent discriminative ability. Classification accuracy by AI ranged from 86% to 98.5%, with corresponding AUC values between 0.873 and 1.0. In studies where AI supported clinicians, diagnostic accuracy rose to 97.1%, compared to unaided human accuracy ranging from 77.5% to 93.5%. The most frequently used convolutional neural network (CNN) architectures were GoogLeNet and DenseNet. Training data proportions varied, with some studies using up to 95% of data for training. Grad-CAM was employed in five studies to visualize model focus areas.