2025-04-212025-04-21https://scholars.lib.ntu.edu.tw/handle/123456789/728381We propose a novel approach for PDAC detection in clinical applications, leveraging the combined power of metabolomics, proteomics, clinical data, biochemical biomarkers, and machine learning. This hybrid AI model will integrate diverse information layers, including metabolic and proteomic profiles, patient clinical parameters (age, gender, smoking status, BMI, HbA1c, or GluAC), and established PDAC markers, to achieve superior diagnostic accuracy and sensitivity compared to traditional methods. This integrated approach can potentially revolutionize PDAC detection, leading to earlier intervention and improved patient outcomes. Furthermore, the scientific findings of the project can undergo cross-validation in both Taiwan and Lithuania, suggesting their potential applicability worldwide to benefit diverse populations. Enhanced international cooperation among medical institutes could facilitate the validation of these results.代謝體蛋白質體人工智慧胰臟癌MetabolomicsProteomicsArtificial IntelligencePancreatic Cancer