When developing tools for automated cortical segmentation, the ability to produce topologically correct segmentations is important in order to compute geometrically valid morphometry measures. In practice, accurate cortical segmentation is challenged by image artifacts and the highly convoluted anatomy of the cortex itself. To address this, we propose a novel deep learning-based cortical segmentation method in which prior knowledge about the geometry of the cortex is incorporated into the network during the training process. We design a loss function which uses the theory of Laplace's equation applied to the cortex to locally penalize unresolved boundaries between tightly folded sulci. Using an ex vivo MRI dataset of human medial temporal lobe specimens, we demonstrate that our approach outperforms baseline segmentation networks, both quantitatively and qualitatively.
@article{arxiv.2303.00795,
title = {Improved Segmentation of Deep Sulci in Cortical Gray Matter Using a Deep Learning Framework Incorporating Laplace's Equation},
author = {Sadhana Ravikumar and Ranjit Ittyerah and Sydney Lim and Long Xie and Sandhitsu Das and Pulkit Khandelwal and Laura E. M. Wisse and Madigan L. Bedard and John L. Robinson and Terry Schuck and Murray Grossman and John Q. Trojanowski and Edward B. Lee and M. Dylan Tisdall and Karthik Prabhakaran and John A. Detre and David J. Irwin and Winifred Trotman and Gabor Mizsei and Emilio Artacho-Pérula and Maria Mercedes Iñiguez de Onzono Martin and Maria del Mar Arroyo Jiménez and Monica Muñoz and Francisco Javier Molina Romero and Maria del Pilar Marcos Rabal and Sandra Cebada-Sánchez and José Carlos Delgado González and Carlos de la Rosa-Prieto and Marta Córcoles Parada and David A. Wolk and Ricardo Insausti and Paul A. Yushkevich},
journal= {arXiv preprint arXiv:2303.00795},
year = {2023}
}
Comments
Accepted at the 28th biennial international conference on Information Processing in Medical Imaging (IPMI 2023)