AI Science Tracker
EN

Chifeng Municipal Hospital

1 discovery from 2021

Discoveries

Histology and clinical features classify recorded HER2-positive recurrence in a separate cohort

Source check: Source checked. Not independent scientific confirmation.

Finding: A ResNet50-based model combined histology features with clinical data to classify recorded recurrence or metastasis in HER2-positive breast cancer. Developed in 127 patients, it achieved a reported AUC of 0.72 in 123 TCGA patients. The result is retrospective clinical validation, without a time-calibrated recurrence forecast or demonstrated treatment benefit.

Technology: ResNet50; multimodal compact bilinear fusion

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

AI contribution
Learned image features and multimodal fusion enable the source-specific classification of recorded recurrence outcomes.
Test and result
ResNet50 patch features were fused with clinical variables; patch scores were averaged per slide. Separate TCGA data supplied the represented AUC.
Limitation / still unknown
Internal two-fold splitting is described in terms of patches/samples without established patient separation. Follow-up horizons and censoring are not adequately specified for a time-calibrated prognosis. Histology heterogeneity and different cohort characteristics limit reproducibility. No prospective validation, avoided treatment or improved clinical outcomes were measured.
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/PMC8733169/ · https://doi.org/10.1016/j.csbj.2021.12.028

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: Primary front, clinical labels/cohort, computational imaging/fusion/classification Methods, recurrence Results and limitations.

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 2/5 · 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 the interactive profile for the timeline, or browse all organizations and model families.