• bioinformatics
• machine learning
• pancreatic cancer
• statistical learning
• surgery

This thesis has been conceived as the embodiment of the emerging choral strategy of multidisciplinarity in the scientific world: a surgical condition, the pancreatic adenocarcinoma, has been analyzed under the lens of the Surgeon, the Engineer and the Biotechnologist.

In particular, a diagnostic tool has been designed to detect pancreatic cancer early and efficiently by deciphering its tumoral genetic expression. A Statistical Learning approach, arising from the fields of Biostatistics and Machine Learning, has been realized step by step in each Chapter, covering the domains of Physics, Mathematics, Biology and Bioinformatics, from data analysis and pre-processing to a thorough model training and assessment.

An experimental original demonstration of the tool’s great potential has concluded the dissertation, by proving that the algorithm has the unique capacity of quadrupling the 5-year survival rate.

Hence, a new paradigmatic mindset in Medicine has been sealed: mastering and controlling the languages of Biotechnologies, Engineering and Surgery as apex of medical education.