LAUSR

Accurate segmentation of the hippocampus for hippocampal avoidance whole brain radiotherapy currently requires high resolution MRI in addition to neuroanatomic expertise for manual segmentation. Removing the need for MR images to identify the hippocampus would reduce planning complexity, the need for a treatment planning MR imaging session, potential uncertainties associated with MRI-CT image registration, and cost. Three-dimensional deep convolutional network models have the potential to automate hippocampal segmentation. In this study, we demonstrate that deep learning models, utilizing 3D convolutional neural networks, can accurately delineate the hippocampus using only high-resolution non-contrast CT images. PURPOSE To investigate the inter-observer accuracy and reliability of hippocampal segmentation by experts using MRI -fusion and an automated deep learning model using CT alone. METHODS Retrospectively, 390 Gamma Knife patients with high resolution CT and MR images were collected. Following the RTOG 0933 guidelines, images were rigidly fused, and a neuroanatomic expert contoured the hippocampus on the MR, then transferred the contours to CT. Using a calculated cranial centroid, the image volumes were cropped to 200 x 200 x 35 voxels, which were used to train four models, including our proposed Attention-Gated 3D ResNet (AG-3D ResNet). These models were then compared with results from a nested 10-fold validation. From the predicted test set volumes, we calculated the 100% Hausdorff distance (HD). Acceptability was assessed using the RTOG 0933 protocol criteria, contours were considered passing with HD ≤ 7 mm. RESULTS The bilateral hippocampus passing rate across all 90 models trained in the nested cross-fold validation was 80.2% for AG-3D ResNet, which performs with a comparable pass rate (p = 0.3345) to physicians during centralized review for the RTOG 0933 Phase II clinical trial. CONCLUSION Our proposed AG-3D ResNet's segmentation of the hippocampus from non-contrast CT images alone are comparable to those obtained by participating physicians from the RTOG 0933 Phase II clinical trial.