Translational challenges of biomedical machine learning solutions in clinical and laboratory settings

The ever increasing use of AI methods in biomedical sciences calls for closer inter-disciplinary collaborations that transfer the domain knowledge from life scientists to IT researchers and vice-versa. We highlight two general areas where the use of AI-based solutions designed for clinical and laboratory settings has proven problematic. These are used to demonstrate common sources of translational challenges that often stem from the differences in data interpretation between the clinical and research view, and the unmatched expectations and requirements on the result quality metrics. To conclude, we outline how explicit interpretable inference reporting might be used as a guide to overcome such translational challenges.The following work has not been published yet and remains as pre-print. In particular, this extended abstract has gone through double-blind peer-review (2 reviewers) to accept a poster contribution at NLDL 2022 however it has not been included in any proceedings or journal. Note from NLDL: Authors of accepted extended abstracts (non-archival submissions) retain full copyright of their work, and acceptance of such a submission does not preclude publication of the same material in another journal or conference