A neoteric ensemble deep learning network for musculoskeletal disorder classification

Abstract

The healthcare area is entirely different from other industries. It is of the highly significant area and people supposed to gain the utmost care and facilities irrespective of the cost. Reliable image detection and classification is considered a significant capability in medical image investigation problems. The key challenge is that the whole image has to be searched for a particular event and then classified accordingly but it is necessary to ensure that any important piece of information or instance shouldn’t be skipped. With regards to image analysis by radiologists, it is quite restricted because of its partiality, the intricacy of the images, wide variations that happen amongst various analysts and weariness. However, the introduction of deep learning is a promising way to improve this situation by sorting out the issue according to human leaning mechanism consequently it brings high-tech changes in medical image classification problems. In this context, a new ensemble deep learning topology is being proposed in the direction of a more precise classification of musculoskeletal ailments. In this regard, a comparison has been accomplished based on different learning rates, drop-out rates, and optimizers. This comparative research proved to be a baseline to gauge the up-to-the-mark performance of the proposed ensemble deep learning architecture.