Classification of X-ray Images for Human Body Parts
1*Tejaswini Reddy Naini, M. Jaiganesh, S. Suguna Mallika, Suchith Buddha
Due to advances in medical imaging technology, there is a proliferation of diagnostic images acquired in medical centres that need to be stored, analysed, retrieved and classified. The development of automatic analysis of X-ray images and classification methods is a pressing need that will have a critical impact on clinical practices by reducing human errors. Analysis of X-ray images is mostly being done by medical specialists, as it is a critical sector and people anticipate the highest level of care and service regardless of cost. Depending on just one technique to gain a high accuracy rate for every individual class is unreliable. In this paper, the classification of medical X-ray images against body parts using a pre-trained deep convolutional neural network (DCNN) and two handcrafted descriptors in a joint approach is enclosed.
Body-parts classification, handcrafted features, deep features, pre-trained CNN, joint approach.