KOA-NN: A convolutional neural network model based on spatial and channel attention mechanisms for detecting knee osteoarthritis from X-ray images

Abstract

Knee osteoarthritis (KOA) is a deformity that causes mobility problems, pain, and wear and tear in the knee joints. Currently, one way to diagnose KOA is through the use of X-ray images. An expert analyzes the images and infers the patient's condition. However, this is a tiring and error-prone task, since the expert must analyze a large number of images to make a diagnosis. For this reason, in this work, a new convolutional neural network (CNN) model based on attention mechanisms called KOA-NN is used to identify KOA accurately. The results of this work position this new model as an alternative to the models established in the literature.