[2604.07180] Energy-based Tissue Manifolds for Longitudinal Multiparametric MRI Analysis

Computer Science > Computer Vision and Pattern Recognition arXiv:2604.07180 (cs) [Submitted on 8 Apr 2026] Title:Energy-based Tissue Manifolds for Longitudinal Multiparametric MRI Analysis Authors:Kartikay Tehlan, Lukas Förner, Nico Schmutzenhofer, Michael Frühwald, Matthias Wagner, Nassir Navab, Thomas Wendler View a PDF of the paper titled Energy-based Tissue Manifolds for Longitudinal Multiparametric MRI Analysis, by Kartikay Tehlan and 6 other authors View PDF HTML (experimental) Abstract:We propose a geometric framework for longitudinal multi-parametric MRI analysis based on patient-specific energy modelling in sequence space. Rather than operating on images with spatial networks, each voxel is represented by its multi-sequence intensity vector ($T1$, $T1c$, $T2$, FLAIR, ADC), and a compact implicit neural representation is trained via denoising score matching to learn an energy function $E_{\theta}(\mathbf{u})$ over $\mathbb{R}^d$ from a single baseline scan. The learned energy landscape provides a differential-geometric description of tissue regimes without segmentation labels. Local minima define tissue basins, gradient magnitude reflects proximity to regime boundaries, and Laplacian curvature characterises local constraint structure. Importantly, this baseline energy manifold is treated as a fixed geometric reference: it encodes the set of contrast combinations observed at diagnosis and is not retrained at follow-up. Longitudinal assessment is therefore formulated...