Ensemble learning-based abnormality diagnosis in wrist skeleton radiographs using densenet variants voting

Almost one out of five people, including children, suffers from musculoskeletal disorders. It is the second leading cause disability worldwide. affects system’s major areas, represented by shoulder, forearm, and wrist. causes severe pain, joint noises, disability. To detect abnormality, radiologist analyzes patient’s anatomy through X-rays different views projections. automatically diagnose abnormality in system a challenging task. Previously, various researchers detected radiographic images using several deep learning techniques. They used capsule network, 169-layer convolutional neural group normalized network detection. However, to propose methods for improving detection, further work needs be done because accuracy conventional far away 90%. This paper presents an ensemble learning-based classification detecting wrist radiographs. Tags radiographs may result noisy features hence reducing performance. Therefore, tags are segmented removed UNet trained on annotated ground truths. Segmented then voting-based diagnosis. The simulation results show that proposed methodology improves testing 1.5%-4.5% compared available detection methods. can any kind