DASS Good: Explainable Data Mining of Spatial Cohort Data

Developing applicable clinical machine learning models is a difficult task when the data includes spatial information, for example, radiation dose distributions across adjacent organs at risk. We describe co-design of modeling system, DASS, to support hybrid human-machine development and validation predictive estimating long-term toxicities related radiotherapy doses in head neck cancer patients. Developed collaboration with domain experts oncology mining, DASS incorporates human-in-the-loop visual steering, data, explainable AI augment knowledge automatic mining. demonstrate two practical stratification report feedback from experts. Finally, we design lessons learned this collaborative experience.