A survey of bone abnormalities detection using machine learning algorithms.

Bone disease means any of the diseases or injuries that affect human bones. Diseases and injuries of bones are major causes of abnormalities of the human skeletal system. Although physical injury, causing fracture, dominates over disease, but fracture is one of several common causes of bone diseases. Bone fracture disease can indirectly lead to death, as fractures and their associated complications can in some cases trigger a downward spiral in health. Approximately 20 percent of the hip fracture patients died within a year of the fracture. The aim of the study was to conduct a systematic literature review of bone abnormalities that used machine learning or Deep Learning as a mean of detection and classification of bone abnormalities, by systematic review technique which included study reference, methods, dataset, programming language used, and the best accuracy obtained. 17 studies published between 2002 and 2020 were selected for analysis. Seven of which uses X-ray images, five uses CT images and five used MRI Images. The results showed that among the classification techniques examined, artificial neural networks, Convolutional Neural Network and Residual neural network widely used with average accuracy values between 73.42% and 94.70%. Most of the selected studies used X-ray images. The result is that it is possible in the future develop a deep learning model to detect abnormalities bone with more accurate results, by modifying the techniques used and adding more effective algorithms. [ABSTRACT FROM AUTHOR]

Copyright of AIP Conference Proceedings is the property of American Institute of Physics and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)