Deep learning for predicting pituitary neuroendocrine tumour lineage and high-risk subtypes from histology
AI analysis of routine tumor slides predicts pituitary cancer subtypes with high accuracy, revealing immune patterns linked to recurrence.
A deep learning model trained on 925 H&E slides and externally validated in two independent cohorts (n=419) achieves AUC 0.912 for pituitary tumor lineage classification directly from routine histology. Spatial transcriptomics analysis of recurrence tumors reveals increased M2 macrophages and decreased CD8+ T cell infiltration, providing mechanistic insight into recurrence biology.
What the study was
- Study design
- Deep learning diagnostic model development and external validation (multicenter)
- Population
- Pituitary neuroendocrine tumor patients
- Sample size
- 1344
- Category
- Diagnostics
- Maturity
- Validated
- Journal
- NPJ Precision Oncology
Why it surfaced
Multicenter validated DL model (n=1344) for pituitary tumor subtype classification from H&E; strong external validation AUCs and spatial biology mechanistic support for recurrence prediction.
A plain-language summary of published research — not medical advice. Talk to a clinician about your own care.