Identifying the optimal deep learning architecture and parameters for automatic beam aperture definition in 3D radiotherapy.

Skylar S GayKelly D KislingBrian M AndersonLifei ZhangDong Joo RheeCallistus NguyenTucker J Netherton Jinzhong YangKristy BrockAnuja JhingranHannah SimondsAnn KloppBeth M BeadleLaurence E CourtCarlos E Cardenas

Published in: Journal of applied clinical medical physics (2023)

DeepLabv3+ and D-LinkNet are most robust to initial hyperparameter selection. Learning rate, nonlinear activation function, and kernel size are also important hyperparameters for improving performance.

Keyphrases

deep learning
artificial intelligence
early stage
convolutional neural network
machine learning
locally advanced
radiation therapy
radiation induced
rectal cancer
electron microscopy