Meeting Banner
Abstract #0700

Convolutional neural network based segmentation of the the spinal cord and intramedullary injury in acute blunt spinal cord trauma

David B McCoy1,2, Sara M Dupont1, Charley Gros3, Jared Narvid1, Julien Cohen-Adad3, and Jason F Talbott1,2

1Radiology and Biomedical Imaging, Zuckerberg San Francisco General Hospital and UCSF, San Francisco, CA, United States, 2Brain and Spinal Injury Center, San Francisco, CA, United States, 3Institute of Biomedical Imaging, NeuroPoly Lab, Polytechnique Montreal, Montreal, QC, Canada

This study aims to develop and validate a convolutional neural network for automatic segmentation of the spinal cord (SC) and intramedullary injury in acute blunt SC trauma patients. Using image augmentation of the axial slice cross section and U-net architecture, we were able to achieve a dice coefficient for SC segmentation of 0.92. The same network architecture was also able to identify areas of intramedullary injury. This is the first study to accurately segment the acute blunt trauma SC. Automatic segmentation of the SC in this population makes automatic biomarker analysis and quantitative prognostication of outcomes possible for SC injury.

This abstract and the presentation materials are available to members only; a login is required.

Join Here