Automatic spinal cord segmentation in pediatric MR images
Colline Blanc1,2, Shiva Shahrampour3, Feroze Mohamed3, and Benjamin De Leener1,2,4
1NeuroPoly Lab, Institute of Biomedical Engineering, Polytechnique Montréal, Montreal, QC, Canada, Montreal, QC, Canada, 2Research Center, Ste-Justine Hospital University Centre, Montreal, QC, Canada, Montreal, QC, Canada, 3Thomas Jefferson University, Philadelphia, PA, United States, Philadelphia, PA, United States, 4Department of Computer and Software Engineering, Polytechnique Montréal, Montréal, QC, Canada, Montreal, QC, Canada
Segmentation of the spinal cord is an essential process for the accurate delineation of spinal cord structures. However, it is a long process and automatic segmentation tools are not adapted to segment the pediatric spinal cord. We therefore developed a tool mixing a neural network and a deterministic method to overcome the limitations. We succeeded in obtaining a segmentation with a dice coefficient of 0.86 on a patient with a spinal cord injury (SCI).
This abstract and the presentation materials are available to members only;
a login is required.