Cancer related works

Learning-Based sensitivity analysis and feedback design for drug delivery of mixed therapy of cancer


Objecive: Use the uncertain model to analyze which are the key parameters of the models, the state and the controller which impact the issue of the therapy. Design prediction dashboard for the success/failure as well as the expectation of the needed drugs.

The time structure of the therapy over time.

The success/failure dashboard as a function of the two states that were found to be the most important from the sensitvity analysis.

Dashboard of the drug use expectation as a function of the initial tumor size when no uncertainties on the model’s parameters prevail. Notice that the larger is the initial number of circulaitng lymphocite the larger is the initial tumor size that can be successfully treated.

Dashboard of the drug use expectation as a function of the initial tumor size when 20% uncertainties on the model’s parameters prevail. Notice that the larger is the initial number of circulaitng lymphocite the larger is the initial tumor size that can be successfully treated.

For more details, refer to the [paper].