Development of a novel AI-based risk score for heart valve surgery that outperformed established risk models. The approach serves as a role model how machine learning/artificial intelligence helps to exploit electronic health records for the benefit of patients.

AI-based prediction of allergic reactions to non-steroidal anti-inflammatory drugs (NSAIDs) from genotypes of patients. This model helped for identifying candidate genes that are associated with those allergies.

AI-based prediction of antibiotics resistances from the genotypes of Pseudomonas aeruginosa, a common pathogen in hospital-acquired infections. The results served as the basis for further investigation of advanced antibiotic resistance mechanisms.

Fully automatic quality control system of prints on optical digital media using a multi-method approach that combines computer vision, fuzzy logic, and machine learning.

Machine learning-based prediction of paper quality from a paper mill's complex set of parameters. The resulting model allowed for optimizing the paper production process beyond hitherto available expert knowledge.