TY - JOUR
T1 - Automatic spinal cord segmentation from axial-view MRI slices using CNN with grayscale regularized active contour propagation
AU - Zhang, Xiaoran
AU - Li, Yan
AU - Liu, Yicun
AU - Tang, Shu Xia
AU - Liu, Xiaoguang
AU - Punithakumar, Kumaradevan
AU - Shi, Dawei
N1 - Publisher Copyright:
© 2021 Elsevier Ltd
PY - 2021/5
Y1 - 2021/5
N2 - Accurate positioning of the responsible segment for patients with cervical spondylotic myelopathy (CSM) is clinically important not only to the surgery but also to reduce the incidence of surgical trauma and complications. Spinal cord segmentation is a crucial step in the positioning procedure. This study proposed a fully automated approach for spinal cord segmentation from 2D axial-view MRI slices of patients with CSM. The proposed method was trained and tested using clinical data from 20 CSM patients (359 images) acquired by the Peking University Third Hospital, with ground truth labeled by professional radiologists. The accuracy of the proposed method was evaluated using quantitative measures, the reliability metric as well as visual assessment. The proposed method yielded a Dice coefficient of 87.0%, Hausdorff distance of 9.7 mm, root-mean-square error of 5.9 mm. Higher conformance with ground truth was observed for the proposed method in comparison to the state-of-the-art algorithms. The results are also statistically significant with p-values calculated between state-of-the-art methods and the proposed methods.
AB - Accurate positioning of the responsible segment for patients with cervical spondylotic myelopathy (CSM) is clinically important not only to the surgery but also to reduce the incidence of surgical trauma and complications. Spinal cord segmentation is a crucial step in the positioning procedure. This study proposed a fully automated approach for spinal cord segmentation from 2D axial-view MRI slices of patients with CSM. The proposed method was trained and tested using clinical data from 20 CSM patients (359 images) acquired by the Peking University Third Hospital, with ground truth labeled by professional radiologists. The accuracy of the proposed method was evaluated using quantitative measures, the reliability metric as well as visual assessment. The proposed method yielded a Dice coefficient of 87.0%, Hausdorff distance of 9.7 mm, root-mean-square error of 5.9 mm. Higher conformance with ground truth was observed for the proposed method in comparison to the state-of-the-art algorithms. The results are also statistically significant with p-values calculated between state-of-the-art methods and the proposed methods.
KW - 2D magnet resonance imaging
KW - Cervical spondylotic myelopathy
KW - Convolutional neural network
KW - Level set evolution
KW - Spinal cord segmentation
UR - http://www.scopus.com/inward/record.url?scp=85103270057&partnerID=8YFLogxK
U2 - 10.1016/j.compbiomed.2021.104345
DO - 10.1016/j.compbiomed.2021.104345
M3 - Article
C2 - 33780869
AN - SCOPUS:85103270057
SN - 0010-4825
VL - 132
JO - Computers in Biology and Medicine
JF - Computers in Biology and Medicine
M1 - 104345
ER -