A New Framework for Segmentation of Flexor Tendons within Carpal Tunnel from Magnetic Resonance Images

Carpal tunnel syndrome (CTS) is one of the most common peripheral neuropathies. To evaluate CTS, segmentation of the flexor tendons within the carpal tunnel from magnetic resonance (MR) images is rather important. In this paper, we propose a new segmentation framework for flexor tendons within the carpal tunnel on MR images. The proposed method consists of two stages, including reference tendon contour (RTC) construction and registration-based segmentation. Initially, we segment the flexor tendons from the most proximally cross-sectional image using the watershed algorithm. The segmented tendons then serve as the RTCs, which are applied to the subsequent registration-based segmentation. Each resulting RTC in the current MR image is registered to its distally adjacent image by a rigid transformation and then is refined by the snake deformation. The registration-based segmentation is sequentially applied on each pair of adjacent cross-sectional images along the proximal to distal direction until the most distal image is segmented. Consequently, the flexor tendons in the entire MR volume are automatically segmented. In the experiments, three MR volumes were used to validate the accuracy of the proposed method. The proposed method showed good agreement on tendon segmentations as compared to the manual results by demonstrating an averaged dice similarity coefficient above 0.8. Moreover, we also demonstrated better segmentation results than conventional methods. The proposed automated segmentation method is helpful for quantitatively measuring the flexor tendons, such as their volumes and shape parameters.