Abstract

This paper presents an end-to-end solution for MRI thigh quadricepssegmentation. This is the first attempt that deep learning methods are used forthe MRI thigh segmentation task. We use the state-of-the-art FullyConvolutional Networks with transfer learning approach for the semanticsegmentation of regions of interest in MRI thigh scans. To further improve theperformance of the segmentation, we propose a post-processing technique usingbasic image processing methods. With our proposed method, we have established anew benchmark for MRI thigh quadriceps segmentation with mean JaccardSimilarity Index of 0.9502 and processing time of 0.117 second per image.

 

Quick Read (beta)