CSE - AI Based Medical Image Analysis

New technologies for acquiring large amounts of medical image data give rise to an ever-increasing demand for effective approaches for medical image data analysis tasks. Deep learning (DL) approaches have yielded remarkably powerful solutions for numerous medical imaging applications, largely outperforming traditional image analysis methods. Comparing to natural scene images, medical image analysis faces several different challenges. Commonly, DL methods rely on a great amount of annotated data for model training. While natural scene images are usually 2D, medical images can be 2D, 3D, and even higher dimensional. In particular, 3D medical images are widely used in basic research and clinical practice. Yet, 3D medical image analysis presents big challenges to DL methods. First, 3D medical images are often of very large sizes (e.g., billions of voxels), and thus incur high computation costs. But, current GPUs are of limited memory for implementing 3D DL models. Further, few efficient automatic techniques for annotating 3D images are available. Since in general only trained medical experts can annotate medical images effectively, medical image annotation is a highly costly and labor-intensive process (even for 2D images). Therefore, how to attain sufficient good quality annotated image data for DL model training while significantly reducing annotation efforts of medical experts is a big bottleneck to the successful development and deployment of DL methods for medical imaging applications. We investigate new DL-based annotation-efficient approaches for various medical image analysis tasks (e.g., segmentation, classification, denoising, etc), such as cell, vessel, lymph nodes, bone, and cartilage segmentation in 3D images, liver and bladder cancer detection, etc.