Abstract
The automatic assessment of hippocampus volume is an important tool in the study of several neurodegenerative diseases such as Alzheimer's disease. Specifically, the measurement of hippocampus subfields properties is of great interest since it can show earlier pathological changes in the brain. However, segmentation of these subfields is very difficult due to their complex structure and for the need of high-resolution magnetic resonance images manually labeled. In this work, we present a novel pipeline for automatic hippocampus subfield segmentation based on a deeply supervised convolutional neural network. Results of the proposed method are shown for two available hippocampus subfield delineation protocols. The method has been compared to other state-of-the-art methods showing improved results in terms of accuracy and execution time.
Highlights
The automatic assessment of hippocampus volume is an important tool in the study of several neurodegenerative diseases such as Alzheimer’s disease
Inspired by Generalized Dice Loss (GDL), we propose in this paper the Generalized Jaccard Loss (GDL) (4) which is a variant of Jaccard loss (3) following the same idea to reduce label size dependency: 2 NC
From the architecture point of view, our model is a 3D UNET variant that uses deep supervision and low-resolution feedback to make easier the training process. We found that this variant worked better that the classic UNET
Summary
The automatic assessment of hippocampus volume is an important tool in the study of several neurodegenerative diseases such as Alzheimer’s disease. One of the most well-known methods for HC subfield segmentation is named A SHS17 that uses a multi-atlas approach combined with a similarity-weighted voting and a boosting-based error correction. We proposed a method named HIPS18 that obtained state-of-the-art results in two different delineation protocols (Winterburn and Kulaga-Yoskovitz) with relatively low processing times thanks to the use of a fast multi-atlas label fusion method called OPAL19. These methods have promising results, their automatic measurements are not close enough to manual tracings in some cases[20]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
