Related Experiment Video
Updated: Jul 26, 2025
05:05
Four-Dimensional CT Analysis Using Sequential 3D-3D Registration
Published on: November 23, 2019
8.0K
Measurement of interspinous motion in dynamic cervical radiographs using a deep learning-based segmentation model
Dae-Woong Ham1, Yang-Seon Choi1, Yisack Yoo1
11Department of Orthopedic Surgery, Chung-Ang University Hospital, College of Medicine, Chung-Ang University, Dongjak-gu, Seoul.
Journal of Neurosurgery. Spine
|June 16, 2023
Summary
A new AI algorithm accurately measures interspinous motion (ISM) after ACDF surgery, showing strong agreement with human experts. This deep learning tool aids in evaluating functional fusion status in clinical settings.
Area of Science:
- Spinal Surgery
- Medical Imaging
- Artificial Intelligence
Background:
- Interspinous motion (ISM) is crucial for assessing functional fusion after anterior cervical discectomy and fusion (ACDF).
- Current methods for measuring ISM face challenges with clinical accuracy and ease of use.
Purpose of the Study:
- To evaluate the feasibility of a deep learning-based segmentation model for measuring ISM in ACDF patients.
- To validate a convolutional neural network (CNN) for automated ISM measurement.
Main Methods:
- Retrospective analysis of dynamic cervical radiographs from 106 ACDF patients.
- Training a CNN algorithm on 150 normal cervical radiographs for spinous process segmentation.
- Validation using interrater reliability (ICC), RMSE, and Bland-Altman analysis against human experts.
Main Results:
- The AI algorithm achieved 99.2% accuracy in spinous process detection.
- High interrater reliability (0.88) between AI and human experts for ISM measurement.
- Mean difference in measurement was 0.02 ± 0.68 mm, with 95% limits of agreement from 0.11 to 1.36 mm.
Conclusions:
- A novel CNN-based autosegmentation algorithm demonstrates strong agreement for measuring ISM.
- The AI tool shows potential to assist clinicians in evaluating post-ACDF segmental motion.
- This technology offers a reliable method for assessing functional fusion status.
