Osteoporosis detection using convolutional neural network based on dual-energy X-ray absorptiometry images

Abstract

Osteoporosis is one of the most common diseases that affect the bones of adults, especially women in menopause, and the reason for this is due to the lack of bone mineral density bone mineral density (BMD). BMD can be measured by X-ray and dual energy X-ray absorptiometry (DEXA) images, this article, focused on using DEXA images for Osteoporosis detection. At first, the original image must passed through the preprocessing stage, during which the noisy parts is reduced, and the useless parts are eliminated, and then the contrast between adjacent areas is increased and the area of interest is allocated. After that, the image is passed in a deep learning model in order to extract the unique features on the basis of which each image is classified. The classification result was excellent with 98% accuracy. The used dataset is “Osteoporosis DEXA scans images” of Spine from Pakistan.

Effect of Changing Targeted Layers of the Deep Dream Technique Using VGG-16 Model

“…CNNs are deep neural models that have been created for handling the image data, and they may be considered as feed-forward ANNs with a variety of the subsampling and convolutional layers [17], [18].…”