Deep Learning for Differentiating Pulmonary Metastasis from Primary Lung Cancer Constructed on Frozen Sections for Intraoperative Diagnosis.
AI model rapidly differentiates lung cancer types from frozen tissue during surgery, supporting real-time surgical decisions under time pressure.
Onozato et al. developed and validated a deep learning model trained on intraoperative frozen section histological images that accurately differentiates pulmonary metastasis from primary lung cancer, achieving high diagnostic accuracy in Modern Pathology. The model addresses a critical time-pressure intraoperative diagnostic scenario, offering AI-assisted decision support when tissue sampling is limited and definitive histological classification is urgently required to guide surgical resection extent.
What the study was
- Study design
- Retrospective validation study of deep learning on intraoperative frozen section histology
- Population
- Thoracic surgery patients with pulmonary lesions undergoing intraoperative frozen section assessment
- Category
- Diagnostics
- Maturity
- Exploratory
- Journal
- Mod Pathol
Why it surfaced
Modern Pathology DL intraoperative pathology tool with direct clinical application in thoracic surgery; validated on human specimens with high journal impact in the AI diagnostics space.
A plain-language summary of published research — not medical advice. Talk to a clinician about your own care.