A Learnable Prior Improves Inverse Tumor Growth Modeling

Biophysical modeling, particularly involving partial differential equations (PDEs), offers significant potential for tailoring disease treatment protocols to individual patients. However, the inverse problem-solving aspect of these models presents a substantial challenge, either due to the high comp...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on medical imaging Vol. 44; no. 3; pp. 1297 - 1307
Main Authors: Weidner, Jonas, Ezhov, Ivan, Balcerak, Michal, Metz, Marie-Christin, Litvinov, Sergey, Kaltenbach, Sebastian, Feiner, Leonhard, Lux, Laurin, Kofler, Florian, Lipkova, Jana, Latz, Jonas, Rueckert, Daniel, Menze, Bjoern, Wiestler, Benedikt
Format: Journal Article
Language:English
Published: United States IEEE 01.03.2025
Subjects:
Algorithms
Biological system modeling
Brain - diagnostic imaging
Brain modeling
Brain Neoplasms - diagnostic imaging
Brain Neoplasms - pathology
CMA-ES
Computational modeling
Deep Learning
evolutionary sampling
Humans
Image Processing, Computer-Assisted - methods
Image segmentation
Individualized brain tumor modeling
inverse biophysics
learnable prior
Magnetic Resonance Imaging - methods
Mathematical models
Models, Biological
MRI
Predictive models
Radiation therapy
Robustness
Training
Tumors
ISSN:0278-0062, 1558-254X, 1558-254X
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Biophysical modeling, particularly involving partial differential equations (PDEs), offers significant potential for tailoring disease treatment protocols to individual patients. However, the inverse problem-solving aspect of these models presents a substantial challenge, either due to the high computational requirements of model-based approaches or the limited robustness of deep learning (DL) methods. We propose a novel framework that leverages the unique strengths of both approaches in a synergistic manner. Our method incorporates a DL ensemble for initial parameter estimation, facilitating efficient downstream evolutionary sampling initialized with this DL-based prior. We showcase the effectiveness of integrating a rapid deep-learning algorithm with a high-precision evolution strategy in estimating brain tumor cell concentrations from magnetic resonance images. The DL-Prior plays a pivotal role, significantly constraining the effective sampling-parameter space. This reduction results in a fivefold convergence acceleration and a Dice-score of 95%.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
ISSN:0278-0062
1558-254X
1558-254X
DOI:10.1109/TMI.2024.3494022