Academia Sinica and National Taiwan University (NTU) Hospital have jointly developed PanMETAI, an AI metabolomics platform for detecting early stages of pancreatic cancer.
The tabular AI model leverages a standardized nuclear magnetic resonance (NMR) analysis platform to detect molecular signatures of cancer.
Discover effective B2B marketing
We combine business intelligence and editorial excellence to reach passionate professionals across 36 major media platforms.
learn more
Pancreatic cancer is often diagnosed at an advanced stage because there are no early symptoms, and the five-year survival rate is only 13%. A new AI metabolomics platform aims to improve accurate screening and patient outcomes by facilitating early diagnosis.
PanMETAI uses a standardized NMR approach to analyze up to 260,000 molecular signals per individual.
Unlike traditional diagnostics that rely on a single biomarker, it captures a comprehensive metabolic profile from precancerous changes to early lesions. This feature addresses a critical gap in risk assessment.
The model underwent validation with an independent blind test in Taiwan and additional evaluation with a European cohort in Lithuania.
Results showed that the area under the curve (AUC) in Taiwan was 0.99, sensitivity was 93%, and specificity was 94%. We also achieved a high AUC of 93% in the Lithuanian cohort.
The development of PanMETAI integrates more than 20 years of clinical experience at NTU Hospital and Academia Sinica research in metabolomics, basic science, and theoretical computational science.
The research team envisions this model as a useful screening tool for high-risk individuals. In the future, this AI system may evolve into a “multi-cancer early prediction platform” and contribute to the advancement of precision medicine.
The study’s lead author is Dan-Ni Wu, a postdoctoral fellow at the Academia Sinica Genomics Research Center.
The corresponding authors are Professor Yu-Ting Chang of NTU Hospital, Distinguished Researcher Chao-Ping Hsu (Institute of Chemistry, Academia Sinica), and Associate Researcher Chun-Mei Hu (Academia Sinica Genomics Research Center).
