Related Experiment Video
Updated: Nov 11, 2025
06:48
Automated Segmentation of Cortical Grey Matter from T1-Weighted MRI Images
Published on: January 7, 2019
9.2K
Automatic spinal cord segmentation from axial-view MRI slices using CNN with grayscale regularized active contour
Xiaoran Zhang1, Yan Li2, Yicun Liu3
1School of Automation, Beijing Institute of Technology, Beijing, 100081, China; Department of Electrical and Computer Engineering, University of California, Los Angeles, Los Angeles, CA, 90095-1594, USA.
Computers in Biology and Medicine
|March 29, 2021
Summary
A new automated method accurately segments the spinal cord in MRI images for cervical spondylotic myelopathy (CSM) patients. This improves surgical precision and reduces complications by precisely identifying the affected spinal cord segment.
Area of Science:
- Medical Imaging
- Neurosurgery
- Artificial Intelligence
Background:
- Accurate spinal cord segmentation is vital for surgical planning in cervical spondylotic myelopathy (CSM).
- Current methods may lack precision, leading to potential surgical trauma and complications.
- Identifying the responsible cervical segment is critical for effective treatment.
Purpose of the Study:
- To develop and validate a fully automated approach for spinal cord segmentation from 2D axial-view MRI slices.
- To enhance the accuracy and reliability of spinal cord segmentation in CSM patients.
- To provide a tool for precise localization of the affected segment, aiding surgical decision-making.
Main Methods:
- A fully automated spinal cord segmentation algorithm was developed.
- The method was trained and tested on 359 MRI slices from 20 CSM patients.
- Ground truth was established by professional radiologists; performance was evaluated using Dice coefficient, Hausdorff distance, and RMSE.
Main Results:
- The proposed automated method achieved a Dice coefficient of 87.0%.
- Quantitative metrics included a Hausdorff distance of 9.7 mm and RMSE of 5.9 mm.
- The method demonstrated higher conformance with ground truth compared to state-of-the-art algorithms, with statistically significant results.
Conclusions:
- The developed automated approach offers accurate and reliable spinal cord segmentation for CSM patients.
- This technique can significantly aid in the precise positioning required for CSM surgery.
- The findings suggest a promising advancement in AI-driven medical image analysis for neurosurgical applications.
