Gaussian mixture model and Bayesian convolutional neural network for abnormal musculoskeletal radiographs classification

Musculoskeletal radiograph classification has been gaining great attention due to its important application in automated diagnosis of musculoskeletal disorders. Generally, musculoskeletal radiographs capture images of various bones in the whole body, which can be considered as a combination of multiple single datasets. Existing works have tried to build either a traditional neural network (NN), convolution neural network (CNN), or Bayesian convolutional neural network (BCNN) to achieve efficient classification. However, none of them explore the mixed-dataset property. In this context, we propose a novel Gaussian mixture model-based BCNN (MixBCNN) for the abnormal musculoskeletal radiographs classification problem. In the proposed method, the distribution of NN weights is represented as a Gaussian mixture distribution, expecting a new neural network model with better representation. In addition, we propose an ensemble learning approach to efficiently combine potential NN backbones. Performance evaluation is examined for the popular MURA dataset and shows the superiority of the proposed model with respect to state-of-the-art BCNN methods, notably by 0.86 in F1 score and 0.72 in Cohen’s kappa metrics while asking for negligible complexity overhead.