| Project Detail |
In Ai4Cilia (“Artificial intelligence for Cilia”), we will develop an automated microfluidic AI-powered assay of ciliated cells to detect and classify ciliary beat abnormalities for drug discovery, biomedicine, and disease screening. Rhythmically beating ciliated cells perform important physiological functions in the airways and the reproductive system. Defects of ciliary beat contribute to debilitating diseases, such as to-date uncurable chronic obstructive pulmonary disease (COPD), and underdiagnosed causes of infertility. In the context of our ERC StG MecCOPD, we recently showed that standard video-microscopy recordings of ciliary beat can be used to extract quantitative metrics that clearly identify disease-specific ciliary dysfunction. Therefore, since ciliary beat can be observed in vitro and in minimally invasive biopsies, it could provide a much-needed readout for drug development and improved diagnosis. However, since ciliary beat patterns are highly complex, it takes specialists to analyze them and hence this information is rarely exploited in praxis. To establish ciliary beat as a common readout, we developed a microfluidic platform that standardizes data acquisition from ciliated cells and uses AI for detecting and classifying ciliary beat defects. To bring this approach to the next level and towards market entry, we will address three key objectives in Ai4Cilia to demonstrate (1) hardware feasibility by validating human ciliated cell capture and data acquisition in our microfluidic chip; (2) software feasibility by training our AI algorithm to detect disease-specific ciliary defects and validate sensitivity and specificity of our assay; and (3) business feasibility by assessing customer needs in biomedical and clinical research to define the most efficient go-to-market strategy. We believe this strategy will make ciliary beat a highly sensitive and robust readout for respiratory and reproductive (dys)function in biomedical and clinical research. |
