Medical data sharing is often impeded by privacy concerns and proprietary interests. Novel methods of generative artificial intelligence (AI) enable the synthesis of medical data that closely mimic properties of real patient data, both in image and tabular form. Synthetic data can be freely shared, tailored to the researchers’ needs, and extrapolated to accommodate for often small sample sizes in rare diseases. In this project, the team develops AI models to generate high-quality synthetic image data from bone marrow aspirates and tabular data for clinical trials in acute myeloid leukemia and multiple myeloma. The aim is to establish synthetic data for training and evaluation of AI classification models and synthetic control cohorts in novel clinical trial designs.
 

Furter Information here.