When can synthetic clinical-trial data support valid statistical inference?
Read a 115-trial oncology stress test of synthetic data, understand false positives and uncertainty propagation, then define fit-for-purpose validation for one ADaM or TFL-testing use case.
115phase 3 randomized oncology trials studied
16generative models evaluated
5minutes of spoken output