CONVOLUTION NEURAL NETWORK COMPARISON ANALYSIS FOR BONE FRACTURE DIAGNOSIS

Identification of bone fractures with computer-aided detection and diagnosis is an utmost need of today. It helps radiologists in saving the time and improving the performance. There were many image processing techniques used earlier for detecting the bone fractures. In the current condition of medical imaging processing, models based on deep learning in specific convolutional neural networks are widely used. It also extends its horizon in bone fracture detection from the X-Ray images. The commonly used dataset for bone fracture detection is MURA Dataset. A computer-aided telemedicine system must include automated fracture identification. Human arbitrary bones frequently fracture as a result of unintentional trauma, such as sliding. In reality, many hospitals are short on qualified surgeons who can identify fractures. Computer-aided diagnosis (CAD) lessens the workload for medical professionals while also detecting fractures. In this study, several deep convolutional neural network models for fracture detection are evaluated. The MURA Dataset has been employed therefore in occurrence. Three models—two variations of the DenseNet169 model and a VGG Model—have been examined.The DenseNet model designed is differentiated from the standard model including the variation in the loss function and the usage of weights. This articlespecifically concentrates on the detection of fracture in humerus bone extracted from the MURA Dataset.