Semi-automatic knee cartilage segmentation using Cellular Automata

• Main Authors: Che-Fu Change, 張哲輔
• Other Authors: Shen, D.F.
• Format: Others
• Language: zh-TW
• Published: 2010
• Online Access: http://ndltd.ncl.edu.tw/handle/30161169410218219734

碩士 === 雲林科技大學 === 電機工程系碩士班 === 98 === Many medical treatment methods are proposed and developed for curing degenerative osteoarthritis in knee, which is a very common disease in the elders. As a joint project with Buddhist Tzu-Chi General Hospital in Ta-Lin, Chia-Yi, we developed a non-invasive diagnostic method to measure the knee cartilage volume based on the magnetic resonance imaging (MRI) sequences. We proposed a semi- automatic method based on the CA (Cell Automata) algorithm for segmentation and volumetric measurement of the knee cartilage; we also implemented an interactive interface for easy and precise operation, which reduces the required operation time by 43% of the manual operation. The precision in terms of reproducibility by the proposed method is measured using the root-mean-square coefficient of variation (RMS CV). Experiments on eight sets of MRI data show that the RMS CV is 1.4% for tibia medial cartilage volume ranging from 0.4 to 2.2, which is among the best compared to RMS CV values ranging from 1.4 to 6.3% by state-of-the art methods. The proposed CA based semi-automatic method is convenient and reliable for knee cartilage measurement.