Meeting Banner
Abstract #2733

Sciatic Nerve Segmentation in MRI Volumes of the Upper Leg via 3D Convolutional Neural Networks

Matthew Hancock1, Shashank Manjunath1, Jun Li2, and Richard Dortch3,4

1Vanderbilt University, Nashville, TN, United States, 2Neurology, Vanderbilt University Medical Center, Nashville, TN, United States, 3Radiology, Vanderbilt University Medical Center, Nashville, TN, United States, 4Biomedical Engineering, Vanderbilt University, Nashville, TN, United States

In Charcot-Marie-Tooth disease (CMT) diseases, sciatic nerve (SN) hypertrophy may be a viable biomarker of patient impairment. Estimating nerve diameters currently requires labor-intensive manual segmentations. Our goal was to use 3D convolutional neural networks (CNN), which have been applied successfully in other biomedical imaging applications, to segment the SN. Using a 3D U-Net architecture developed in Keras 2.0 and Python 2.7, we trained CNNs on data partitioned from 38 control and 34 CMT patients with manually defined region-of-interests (ROI). We found that batch-normalizing 3D CNNs achieved the highest performance, demonstrating CNN’s ability to automatically produce high-quality segmentations of the SN.

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

Join Here