A deep learning model for detection of cervical spinal cord compression in MRI scans

Zamir Merali,Michael G Fehlings,Christopher D Witiw,Justin Z Wang,Jefferson R Wilson,Jetan H Badhiwala

doi:10.1038/s41598-021-89848-3

Zamir Merali, Michael G Fehlings + Show 4 more

Open Access

https://doi.org/10.1038/s41598-021-89848-3

Copy DOI

Abstract

Magnetic Resonance Imaging (MRI) evidence of spinal cord compression plays a central role in the diagnosis of degenerative cervical myelopathy (DCM). There is growing recognition that deep learning models may assist in addressing the increasing volume of medical imaging data and provide initial interpretation of images gathered in a primary-care setting. We aimed to develop and validate a deep learning model for detection of cervical spinal cord compression in MRI scans. Patients undergoing surgery for DCM as a part of the AO Spine CSM-NA or CSM-I prospective cohort studies were included in our study. Patients were divided into a training/validation or holdout dataset. Images were labelled by two specialist physicians. We trained a deep convolutional neural network using images from the training/validation dataset and assessed model performance on the holdout dataset. The training/validation cohort included 201 patients with 6588 images and the holdout dataset included 88 patients with 2991 images. On the holdout dataset the deep learning model achieved an overall AUC of 0.94, sensitivity of 0.88, specificity of 0.89, and f1-score of 0.82. This model could improve the efficiency and objectivity of the interpretation of cervical spine MRI scans.

Highlights

Magnetic Resonance Imaging (MRI) evidence of spinal cord compression plays a central role in the diagnosis of degenerative cervical myelopathy (DCM)
As DCM is a heterogenous condition we aimed to develop a model that would have similar performance in patients with various demographics, disease characteristics and on images gathered with various MRI scanners
This study involved retrospective analysis of prospectively collected magnetic resonance imaging (MRI) studies from patients with DCM enrolled in the AO Spine CSM North America (CSM-NA; ClinicalTrials.gov NCT00285337) or AO Spine CSM International (CSM-I; Clinical Trials.gov NCT00565734)

Summary

Introduction

Magnetic Resonance Imaging (MRI) evidence of spinal cord compression plays a central role in the diagnosis of degenerative cervical myelopathy (DCM). We aimed to develop and validate a deep learning model for detection of cervical spinal cord compression in MRI scans. On the holdout dataset the deep learning model achieved an overall AUC of 0.94, sensitivity of 0.88, specificity of 0.89, and f1-score of 0.82 This model could improve the efficiency and objectivity of the interpretation of cervical spine MRI scans. No studies to date have attempted to use deep learning methods to detect spinal cord compression in a population of patients with DCM. Our aim was to develop a novel deep learning model to detect cervical spinal cord compression in patients with DCM in T2 weighted MRI scans. After developing a model we attempted to gain insights into how the model functioned using analytic techniques

Objectives

Methods

Results

Discussion

Conclusion