MétaCan
Menu
← Back to cohort
Record W4409425345 · doi:10.1101/2025.04.12.25325548

Computational characterization of lymphocyte topology on whole slide images of glomerular diseases

2025· preprint· en· W4409425345 on OpenAlexfundno aff
Xiang Li, Manav Shah, Qian Liu, Jin Zhou, Gina Sotolongo, Jeffrey B. Hodgin, Laura H. Mariani, Lawrence B. Holzman, Andrew Janowczyk, Jarcy Zee, Kyle J. Lafata, Laura Barisoni

Bibliographic record

VenuemedRxiv · 2025
Typepreprint
Languageen
FieldMedicine
TopicSystemic Lupus Erythematosus Research
Canadian institutionsnot available
FundersUniversity of North Carolina at Chapel HillTemple UniversityUniversity of PittsburghAnschutz Medical Campus, University of ColoradoJohns Hopkins UniversityNIH Clinical CenterWashington University in St. LouisUniversity of South CarolinaUniversity of Texas Southwestern Medical CenterUniversity of MinnesotaWake Forest UniversityChildren's Mercy HospitalTexas Children's HospitalHospital for Sick ChildrenNorthwestern UniversityUniversity of WashingtonNational Cancer InstituteEmory UniversityCleveland ClinicUniversity of PennsylvaniaOhio State UniversityYork UniversityUniversity of Miami
KeywordsResamplingCluster analysisProportional hazards modelComputer scienceArtificial intelligenceMultivariate statisticsMedicinePattern recognition (psychology)Machine learningInternal medicine

Abstract

fetched live from OpenAlex

The complexity of distribution of inflammatory cells in the kidney is not well captured by conventional semiquantitative visual assessment. This study aims to computationally quantify the topology of lymphocytic inflammation and tested its clinical relevance. N=333 NEPTUNE/CureGN participants (N=155 focal segmental glomerulosclerosis (FSGS) and N=178 Minimal Change Disease (MCD) with available clinical/demographic data and 1 Hematoxylin & Eosin-stained whole slide image (WSI), were included. Deep learning models were applied to segment cortex and lymphocytes. Graph modeling, where nodes were defined as lymphocytes and edges as the spatial connections between cortical lymphocytes, were applied to all WSIs. We then developed a novel graph-based habitat clustering algorithm to identify dense vs. sparse lymphocytic habitats. From each habitat, 26 high-throughput quantitative pathomic features were extracted to capture cell density, connectivity, clustering, and centrality. The association of these pathomic features with disease progression (40% eGFR decline or kidney replacement therapy) was assessed using LASSO-regularized Cox proportional hazards models. Clinical and demographic characteristics were added as potential confounders. Kaplan-Meier survival analysis with log-rank test was used to evaluate risk stratification. Two validation strategies were applied: (i) training on NEPTUNE with external validation on CureGN data, and (ii) using an 80/20 data partition of the combined datasets for training and validation, respectively. Multivariable Cox models integrating clinical/demographic variables with graph features achieved validation concordance index of 0.736±0.072 in the CureGN external validation and 0.757±0.071 in the combined validation dataset. The average degree feature (overall connectivity) in dense habitat and k-core feature (clustering pattern strength) in sparse habitat revealed consistent association with clinical outcome. The topological characterization of lymphocytic inflammation identifies immune habits, capturing the complexity of pattern of inflammation beyond human vision. These pathomic/topology signatures represent potential digital biomarkers that can enhance our ability to prognosticate/predict clinical outcome in MCD/FSGS.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame machine prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.

metaresearch head score (Codex)0.000
metaresearch head score (Gemma)0.001
Version: metacan-v3-hybrid-931329e0061cValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Simulation or modeling · Consensus signal: Simulation or modeling
GenreCandidate signal: Empirical · Consensus signal: Empirical
Teacher disagreement score0.004
Threshold uncertainty score0.008

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0000.001
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0000.001
Bibliometrics0.0010.001
Science and technology studies0.0000.000
Scholarly communication0.0010.000
Open science0.0000.000
Research integrity0.0000.000
Insufficient payload (model declined to judge)0.0010.000

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.019
GPT teacher head0.315
Teacher spread0.297 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.

The models applied no category: nothing in the taxonomy fit this work.
Study designSimulation or modeling
Domainnot available
GenreEmpirical

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations0
Published2025
Admission routes1
Has abstractyes

Explore more

Same venuemedRxiv→Same topicSystemic Lupus Erythematosus Research→French-language works237,207→