CS Paper Radar

MRI-based Deep Radiomic Phenotyping of Neuromuscular Disorders: A Topology-driven Characterization

Martyna Żur, Łukasz Piórecki, Marek Socha, Jose Verdu Diaz, Ana Topf, Michael Obach, Marco Savarese, Jocelyn Laporte, Benedikt Schoeser, Filippo Santorelli, Jordi Diaz-Manera, Volker Straub, Rossella Tupler, Joanna Polańska

cs.CV ·submitted 2026-08-25 11:30 UTC · v3
unusual

Abstract

Quantitative assessment of muscle MRI is crucial for monitoring neuromuscular disorders (NMD). This study introduces an automated radiomic phenotyping framework based on original features engineered across five main architectural domains: quantitative morphometry, spatial distribution, geometric shape, interactions between progressive fat replacement stages, and graph-based topology. Utilizing 1184 MRI scans from the CoMPaSS-NMD project, we map the complex 3D architecture of heterogeneous intramuscular lipodegeneration into objective, morphologically interpretable biomarkers. We introduce a graph-based skeletonization of fat infiltrates to quantify muscle architectural changes, establishing a multi-dimensional extension of traditional, spatially-agnostic volume metrics by mapping topological networks across the entire 3D muscle volume. Statistical screening via non-parametric Kruskal-Wallis analysis confirmed the discriminative power of these novel descriptors across the genetic hierarchy. Notably, topological network metrics (e.g., SF1_Skel_Nodes, $ε^2$ = 0.2656) and interface dynamics metrics (e.g., SF2_To_SF1_Dist_Min, $ε^2$ = 0.2092) demonstrated substantial effect sizes, providing deeper structural insights than classical volumetric assessments. Post-hoc pairwise evaluations and UMAP projections further indicated the capability of these topological and 3D geometric invariants to capture disease-specific macroscopic infiltration patterns. These results demonstrate that global architectural features represent a highly promising class of biomarkers for differential diagnosis, offering new avenues for tracking longitudinal disease dynamics in neuromuscular diagnostics. The developed automated feature extraction pipeline is integrated and available within the MUSCAT (MUSCle fAt Topology) library.

Comment
Draft manuscript. Has not yet been peer-reviewed

Closest papers

Nearest neighbours by meaning across the whole index.

Unified Framework with Convex-Concave Decomposition and Sparse Topological Learning for Skeleton-Based Clinical Assessment

cs.CV cs.AI ·2026-11-07 ·Quang-Anh N. D., Duc Pham Minh, Minh-Anh Nguyen, Hien Quang Kha, Manh Hung Ha +2 more
similarity 0.60
Abstract

Skeleton-based rehabilitation assessment faces fundamental limitations of fixed topologies which cannot model compensatory patterns, uniform joint treatments overlook clinically important deviations, and lack of geometric understand- ing. We propose Concave-Convex Decomposition Sparse Topological Graph Convolutional Network (CC2-STG), a multi-stream model that combines geo- metric decomposition, sparse feature learning, and adaptive topology. Our framework decomposes joint relationships into complementary convex-concave distance functions to model both local and global movement characteristics, with sparse regularization with proximal optimization selecting clinically relevant geo- metric features, while multi-scale topological analysis captures non-anatomical joint dependencies in pathological movements. Experiments on the KiMoRe and UIPRMD datasets validate the effectiveness of our approach, achieving state-of-the-art performance on clinical pathological movement assessment.

0.23

Predicting Cognitive Profiles from Structural MRI using Spatial Graph Representation Learning

cs.CV ·2026-10-28 ·Taslima Ferdaus Shuva, Md. Tanvir Rahman, Risala Tasin Khan, Mohammad Ali Moni, ASM Ashraf Mahmud
similarity 0.58
Abstract · 15 pages, 4 figures. This work has been submitted to the Elsevier Array for possible publication

Predicting cognitive abilities from structural neuroimaging remains challenging due to the high dimensionality and geometric complexity of brain morphology. Conventional models typically flatten regional morphometry into tabular features, failing to explicitly represent spatial relationships. We propose a Spatial-Topological Graph Neural Network (ST-GNN) to predict four distinct cognitive phenotypes across large pediatric cohorts (N = 2,378 to 2,424). By projecting regional morphometry onto a population-level standardized spatial graph, ST-GNN restricts message passing to defined neighboring regions. Evaluations demonstrate that ST-GNN effectively fuses spatial morphology with systemic covariates, outperforming linear baselines and topology-free deep models. Ablations reveal that imposing a sparse geometric topology (K-nearest neighbors) and utilizing global feature scaling - which preserves absolute morphometric scale - outperform fully connected networks and region-normalized approaches. These choices provide transferable architectural inductive biases, revealing differences between computational routing and morphometric feature attribution. Integrated Gradients analysis identified distinct morphological features consistent with neuroanatomical evidence, such as cerebellar volume for working memory and superior frontal surface area for executive function. These findings suggest standardized structural MRI provides useful information for task-specific cognitive prediction, establishing a foundation for MRI-based neurodevelopmental assessment.

0.28

Targeted Deep Regression - Initial Applications in Whole-Body Water-Fat MRI in UK Biobank

cs.CV cs.LG ·2026-10-26 ·Filip Marttala, Taro Langner, Johan Öfverstedt, Nouman Ahmad, Håkan Ahlström +2 more
similarity 0.58
Abstract · 16 pages, 8 figures

UK Biobank is a British clinical study which, in addition to large amounts of medical data collected, plans to include detailed Magnetic Resonance Imaging (MRI) examina- tions from 100,000 participants. The image data can be analyzed for example by auto- mated analysis with neural networks for image segmentation or classification to expand the collection of image-derived phenotypes for medical research. In this work, we pro- pose, present and evaluate a methodology that uses image-based regression to estimate numerical properties from the collected neck-to-knee water-fat MRI scans. It applies a ResNet50-style network to 2D multi-channel collages that represent targeted and study- specific anatomical sub-regions from the MRI scan. The targeted sub-region sampling is achieved by propagating manually defined subregions from a reference scan to the other >30,000 analyzed scans using image registration. The resulting collages can be compact while retaining high-resolution, for example by providing multiple slices of kidney tissue for a kidney volume regression network. This approach also enables group-wise saliency analysis that highlight image and anatomical regions used by the network for the pre- dictions. The method was applied for the regression endpoints of kidney volume, heart volume, and age. Predictive performance was found between R2 = 0.731-0.984 and the group-wise saliency maps were found to highlight medically relevant regions. We con- clude that the method can be used to standardize and restrict the information presented to a deep regression analysis pipeline, increasing information density and opening up a wide range of future medical inference studies.

0.24

TopoHyperClinicalGraph: A Geometry and Topology-Aware Multimodal Framework for Cancer Survival Prediction

eess.IV cs.CV ·2026-10-30 ·Noaman Khan, Zubair Shah
similarity 0.57
Abstract

We developed TopoHyperClinicalGraph, a multimodal framework combining hyperbolic tissue geometry, persistent topology and clinical graph learning for cancer survival prediction. Pre-extracted UNI2-h features are reduced from 1,536 to 1,024 dimensions using training-fold PCA. A pathology backbone combines hyperbolic neighborhoods with Lorentz-centroid attention pooling. Once frozen, its survival logits are adjusted using persistence descriptors from up to 64 landmarks per slide and sampling controls. A five-node clinical graph then learns a further correction from age, sex, stage, T and N. Five-fold patient-level evaluation covered eight TCGA and CPTAC cohorts with 2,364 patients. Mean validation-selected C-indices for TopoHyperClinicalGraph versus pathology-only spatial PatchGCN were 0.7563 versus 0.7386 in COADREAD, 0.6220 versus 0.5998 in HNSC, 0.6755 versus 0.6535 in BRCA and 0.6409 versus 0.6176 in STAD. CPTAC scores were 0.5984 in HNSC, 0.6060 in PDA, 0.8055 in CCRCC and 0.7192 in LUAD. CPTAC-HNSC used TCGA-initialized adaptation; other CPTAC backbones were trained within their cohorts. Geometry accounted for most pathology-only gains, with smaller persistence effects. Clinical graph fusion exceeded its matched node-wise control in five cohorts, tied in one and scored lower in two; clinical-only prediction exceeded fusion in three CPTAC cohorts. The framework supports separate evaluation of geometric, topological and clinical contributions. All scores were validation-selected; independent testing is needed to establish generalizability.

0.15

Graphlets as structural fingerprints of complex networks

cs.SI q-bio.QM ·2026-09-15 ·Anna Pidnebesna, David Hartman, Aneta Pokorna, Daniel Trlifaj, Jaroslav Hlinka
similarity 0.54
Abstract

Complex networks are often compared using selected graph-theoretical measures that capture a selected set of properties with effects ranging from local to global, such as degree, clustering or betweenness centrality. Here we introduce a structural fingerprinting framework based on graphlets: small rooted subgraphs whose distributions provide a systematic description of local-to-mesoscale topology. Across synthetic networks generated from several random graph models, graphlet fingerprints capture parameter-dependent structural differences, outperform standard graph-theoretical measures, and identify even subtle local patterns driving discrimination. We then apply the framework to empirical resting-state functional connectomes, documenting that while graphlets show superior sensitivity also to controlled topological perturbations of brain connectivity, specifically in schizophrenia-control classification they perform only comparably to classical graph-theoretical features. This is in line with the notion that schizophrenia-related alterations are dominated by spatially localized connectivity changes rather than general topological reorganization. Altogether, the generative modeling, targeted perturbations and real-world neuroimaging classification challenge position graphlets as flexible structural fingerprints of complex networks, while carefully outlining their strength and weaknesses compared to more classical graph theoretical features.

0.16

Target-Aware State-Adaptive $p$-Dirichlet Graph Neural Regression for Non-Invasive Body-Composition Estimation

cs.LG ·2026-08-30 ·Nadejda Drenska, Matthew Lemoine, Gowri Priya Sunkara, Yu Wang, Sri Lakshmi Sravani Devarakonda +1 more
similarity 0.54
Abstract · 21 pages, 7 figures, 3 tables. The first four authors contribute equally to this work and are listed alphabetically. Corresponding authors: …

Accurate estimation of body-composition outcomes, including body fat percentage (BFP), bone mineral density (BMD), and appendicular lean mass (ALM), is important for evaluating metabolic, skeletal, and muscular health. Direct assessment using dual-energy X-ray absorptiometry (DXA), however, requires specialized equipment and involves ionizing radiation. We propose a target-aware, state-adaptive $p$-Dirichlet energy-flow graph neural regression ($p$SADE-GNR) framework for estimating these outcomes from non-invasive anthropometric measurements. A neural encoder maps participant representations to hidden states that are propagated over an outcome-specific participant-similarity graph by a state-adaptive forward-Euler discretization of the graph $p$-Dirichlet energy flow. Graph distances weight each original or latent coordinate by its normalized absolute training-fold correlation with the outcome. Using clinical data from the Pennington Biomedical Research Center and five-fold cross-validation, the correlation-weighted model using the original standardized measurements achieved the lowest root mean squared error in all nine primary outcome-cohort combinations and outperformed previously reported support vector regression or least-squares support vector regression reference values in eight of nine comparisons. Autoencoder, variational-autoencoder, and Gaussian-mixture variational-autoencoder representations generally did not improve primary-outcome prediction or reduce computational cost. In an exploratory age-prediction analysis including ALM, BMD, and BFP as predictors, the correlation-weighted GMVAE model achieved the lowest mean error in all three cohorts. These results support target-aware, state-adaptive $p$-Dirichlet graph neural regression for non-invasive body-composition estimation.

0.29

Predicting Transmembrane Protein Topology from 3D Structure

cs.AI ·2026-09-24 ·Sitong Chen, Xiaopeng Mao
similarity 0.53
Abstract

This paper presents a novel approach to infer protein topology using the state-of-the-art graph neural network (GNN), SchNet. The model is trained on the same dataset used to develop the recent DeepTMHMM model with 5-fold cross-validation. Unlike the conventional approaches based on using only the protein sequences or the $α$-carbons as features, we have decoded our classifier in this way, so all atom-level embeddings are used. Without applying any pre-trained weight, the final results have shown great potential that GNNs can be used for topological predictions.

0.28

CellPath-Bench: A Multidimensional Benchmark for Whole-Slide Cellular Representations in Pathology Foundation Models

cs.AI cs.CV ·2026-08-21 ·Bokai Zhao, Yiyang Zhang, Hanqing Chao, Yawei Ma, Long Bai +4 more
benchmark similarity 0.53
Abstract

Pathology foundation models (PFMs) are increasingly used as general-purpose backbones, yet existing benchmarks cannot systematically diagnose their whole-slide cellular representation capabilities, including the decodability of cell-type information and the transferability of such information across tissue sections, datasets, and anatomical organs. We introduce CellPath-Bench, a cellular-resolution benchmark that evaluates frozen PFMs themselves. Following quality control of 52 candidate Xenium datasets, we construct a panel of 25 spatially aligned H\&E--Xenium tissue sections spanning 11 organs and 7,079,283 cells, harmonized into fine- and coarse-grained taxonomies. CellPath-Bench samples frozen WSI feature maps at registered nuclear coordinates and evaluates them using standardized multiclass linear probes. Cell Representation Advantage (CRA) measures the within-section advantage of nucleus-anchored representations over patch-level mean pooling, while Cell Representation Transferability (CRT) characterizes the generalization of cell-type decodability across tissue sections, datasets, and organs. We benchmark 30 pathology-specific and general-purpose foundation models through 304,920 runs across spatial readouts, magnifications, taxonomic granularities, and evaluation protocols. The results reveal substantial model-dependent differences in cell-type decodability and its cross-domain generalization, yielding distinct multidimensional capability profiles. CellPath-Bench provides a standardized framework for auditing cellular information in frozen PFM representations.

0.15