Bone Fracture Detection using CNN and VGG

Technology has played a huge role in human evolution since time immemorial. Now it is not only limited to engineers and data scientists, but far beyond, to even medicine. With increasing load on medical professionals and healthcare systems, digitalization can come handy. The traditional X-ray imaging gives unclear images of bone components in tissue making it difficult for surgeons to diagnose. Using technology, accurate diagnosis can be achieved. This includes a series of steps comprising pre-processing, edge detection, feature extraction and machine learning classifications. Machine learning algorithms such as Convolutional Neural Networks(CNN) and Visual Geometry Group(VGG) were used on a dataset consisting of 9246 training, 506 test and 829 validation images with 0.95 and 0.85 accuracies respectively. The data loss for CNN was found to be 0.24 in comparison to 0.4 for VGG. So from statistical inference, CNN is better.