The abstract is read at the publisher; the summary is JClub's.
CTransNet, a multitask hybrid ResNet50–Transformer model, classifies tertiary lymphoid structure status in gastric cancer with AUCs up to 0·953 for TLS2.
- Why it matters: Peritoneal recurrence after curative gastric cancer surgery is common and fatal, and histology-based TLS assessment is invasive and biased, limiting patient stratification.
- What they did: Using retrospective data from eight Chinese cohorts (training, internal/external validation, plus two anti-PD-1/anti-PD-L1 cohorts), the model predicted TLS0–TLS2 and peritoneal recurrence-free survival, then stratified risk by the training cohort median score.
- The result: Low predicted scores were linked to higher 5-year peritoneal recurrence rates than high scores (66·69% vs 26·28% in the training cohort), and high scores showed higher immunotherapy objective response rates (54·5% vs 5·6% and 53·1% vs 9·7%).
The findingWhy it mattersWhat they didThe result
AI summary of the abstract · may contain errors · not medical advice · read the paper.