No abstract

Deep learning technique has made a tremendous impact on medical image processing and analysis. Typically, the procedure of medical image processing and analysis via deep learning technique includes image segmentation, image enhancement, and classification or regression. A challenge for supervised deep learning frequently mentioned is the lack of annotated training data. In this paper, we aim to address the problems of training transferred deep neural networks with limited amount of annotated data. We proposed a versatile framework for medical image processing and analysis via deep active learning technique. The framework includes (1) applying deep active learning approach to segment specific regions of interest (RoIs) from raw medical image by using annotated data as few as possible; (2) generative adversarial Network is employed to enhance contrast, sharpness, and brightness of segmented RoIs; (3) Paced Transfer Learning (PTL) strategy which means fine-tuning layers in deep neural networks from top to bottom step by step to perform medical image classification or regression tasks. In addition, in order to understand the necessity of deep-learning-based medical image processing tasks and provide clues for clinical usage, class active map (CAM) is employed in our framework to visualize the feature maps. To illustrate the effectiveness of the proposed framework, we apply our framework to the bone age assessment (BAA) task using RSNA dataset and achieve the state-of-the-art performance. Experimental results indicate that the proposed framework can be effectively applied to medical image analysis task.

Segmentation of Femur bone from X-ray images is an indispensable step in computer aided analysis of medical images and orthopaedic examinations. It is more complex than segmentation from CT and MR images, due to some associated less dense tissues that are hard to distinguish from the femur bone in X-ray images. This paper presents an improved method based on U-Net to automatically extract the femurs from hip X-ray images. This method changes the structure of the U-Net network, which can effectively map the non-linear relationship between hip image and femur image, and accurately segment femur image. The paper also added the absolute deviation loss function to improve the segmentation effect. Experimental results show that this method is accurate, robust, and achieves an average dice similarity coefficient of 0.966. The segmentation results are satisfactory.

No abstract

No abstract