AI Science Tracker
EN

Learned cell maps reveal opposing spatial-density associations with breast-cancer survival

UT Southwestern Medical Center; Northwestern University · 24 Apr 2023 · medicine

Discovery

Learned cell maps reveal opposing spatial-density associations with breast-cancer survival

Source check: Source checked. Not independent scientific confirmation.

Finding: An analysis of breast-cancer tissue images linked two tumor–lymphocyte spatial-density features with survival in opposite directions. Neural nucleus classification enabled the measurements in 822 selected TCGA cases. The retrospective associations and within-dataset prognosis evaluation do not establish immune mechanisms, treatment response or clinical benefit.

Technology: HD-Yolo nucleus detection, classification and segmentation

Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.

AI contribution
A learned nucleus detector/segmenter classified cell types, enabling spatial-density measurements subsequently analyzed using ordinary survival models.
Test and result
The study used 653 infiltrating-ductal and 169 infiltrating-lobular female breast-cancer cases with stage information. Patients were split into 658 training and 164 testing cases. Neural cell labels supported tumor, stromal and lymphocyte density maps; ElasticNet-regularized Cox models combined image and clinical features. Lymphocyte-to-tumor spatial projection had a negative model coefficient, while tumor-to-lymphocyte projection had a positive coefficient; training selection and test survival ranking were described.
Limitation / still unknown
Feature/model selection and association estimates are exploratory within one public cohort. No lobular-only test analysis was possible because that subgroup had no deaths in the test set. Opposing spatial coefficients are model associations and do not prove lymphocyte suppression or tumor inhibition. Statements that the features were more prognostically significant than receptor status are not independent evidence of clinical superiority. No external survival replication, therapy-effect study or clinical decision benefit was established.
Evidence sources and annotation context

Evidence context (descriptions follow the cited sources; test type is a catalog classification):

Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC11161202/

Source checked · editorial details

The source page was checked against this entry. Independent confirmation is not assessed by this label.

Source type: paper. Paper: linked separately.

Review date: 21 Nov 2026. Review scope: Original clinical-cohort, neural-pipeline, survival-selection Methods and spatial-feature Results; model interpretation kept separate from immune causation..

No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.

Editorial assessment: Impact 1/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.

Open in the full catalog for filters, charts and related discoveries. Browse medicine discoveries.