PhysNet embeds physical tumor growth dynamics directly into the latent feature space of a CNN, rather than just as a constraint on the output.
March 31, 2026
Original Paper
Physics-Embedded Feature Learning for AI in Medical Imaging
arXiv · 2603.28057
The Takeaway
By integrating reaction-diffusion models into intermediate layers, the model learns biologically meaningful parameters (like diffusion rates) alongside classification features. This provides a level of interpretability and physical consistency that standard 'black box' medical imaging models lack.
From the abstract
Deep learning (DL) models have achieved strong performance in an intelligence healthcare setting, yet most existing approaches operate as black boxes and ignore the physical processes that govern tumor growth, limiting interpretability, robustness, and clinical trust. To address this limitation, we propose PhysNet, a physics-embedded DL framework that integrates tumor growth dynamics directly into the feature learning process of a convolutional neural network (CNN). Unlike conventional physics-i