The intriguing effect of frequency disentangled learning on medical image segmentation

Guanghui Fu; Gabriel Jimenez; Sophie Loizillon; Lydia Chougar; Didier Dormont; Romain Valabregue; Ninon Burgos; Stéphane Lehéricy; Daniel Racoceanu; Olivier Colliot

doi:10.1117/12.2692286

2 April 2024 The intriguing effect of frequency disentangled learning on medical image segmentation

Guanghui Fu, Gabriel Jimenez, Sophie Loizillon, Lydia Chougar, Didier Dormont, Romain Valabregue, Ninon Burgos, Stéphane Lehéricy, Daniel Racoceanu, Olivier Colliot

Author Affiliations +

Proceedings Volume 12926, Medical Imaging 2024: Image Processing; 129261H (2024) https://doi.org/10.1117/12.2692286
Event: SPIE Medical Imaging, 2024, San Diego, California, United States

Abstract

Deep models have been shown to tend to fit the target function from low to high frequencies (a phenomenon called the frequency principle of deep learning). One may hypothesize that such property can be leveraged for better training of deep learning models, in particular for segmentation tasks where annotated datasets are often small. In this paper, we exploit this property to propose a new training method based on frequency-domain disentanglement. It consists of three main stages. First, it disentangles the image into high- and low-frequency components. Then, the segmentation network model learns them separately (the approach is general and can use any segmentation network as backbone). Finally, feature fusion is performed to complete the downstream task. The method was applied to the segmentation of the red and dentate nuclei in Quantitative Susceptibility Mapping (QSM) data and to three tasks of the Medical Segmentation Decathlon (MSD) challenge under different training sample sizes. For segmenting the red and dentate nuclei and the heart, the proposed approach resulted in considerable improvements over the baseline (respectively between 8 and 16 points of Dice and between 5 and 8 points). On the other hand, there was no improvement for the spleen and the hippocampus. We believe that these intriguing results, which echo theoretical work on the frequency principle of deep learning, are of interest for discussion at the conference. The source code is publicly available at: https://github.com/GuanghuiFU/frequency_disentangled_learning.

Citation Download Citation

Guanghui Fu, Gabriel Jimenez, Sophie Loizillon, Lydia Chougar, Didier Dormont, Romain Valabregue, Ninon Burgos, Stéphane Lehéricy, Daniel Racoceanu, and Olivier Colliot "The intriguing effect of frequency disentangled learning on medical image segmentation", Proc. SPIE 12926, Medical Imaging 2024: Image Processing, 129261H (2 April 2024); https://doi.org/10.1117/12.2692286

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available

Members: $17.00

Non-members: $21.00 ADD TO CART

PROCEEDINGS
8 PAGES

DOWNLOAD PAPER SAVE TO MY LIBRARY

GET CITATION

RIGHTS & PERMISSIONS

Get copyright permission Get copyright permission on Copyright Marketplace

KEYWORDS

Image segmentation

Education and training

Feature fusion

Deep learning

Medical imaging

Heart

Data modeling

Show All Keywords

Keywords/Phrases

Search In:

Publication Years