In this GitHub repository I have made three AI models to help optimize clinical trials. The first model labelled Clinical_Trial_Outcome_Prediction.ipynb utilizes the Pytrial librabry to output a result from 0 to 1 indicating whether a clinical trial will fail or succeed. The second modelPatient_Outcome_Prediction is also a result from 0 to 1 and tells us how a patient will react to a treatment administered during a clinical trial. Lastly, Trial_Similarity_Search.ipynb matches a clinical trial with five other trial that are similar and this can help with trial design.
For more information here's an article I wrote on how you can also implement AI in clinical trial applications: https://medium.com/@ali.muhammad.usman08/how-ai-will-revolutionize-the-process-of-clinical-trials-332e73434b8d.
For more detailed steps on documentation check out https://pytrial.readthedocs.io/en/latest/ and https://github.com/RyanWangZf/PyTrial.