1 h 1 min
Seminar, Hosted by Fields Institute for Research in Mathematical Sciences
Smart Data, Smarter Models: Enhancing the Predictive Power of Mathematical Models of Cancer
The College of New Jersey
Recording
Abstract
Mathematical models can vastly improve understanding of cancer dynamics and treatment response, but to be useful they need experimental or clinical data to train and validate them, and not all data are created equal. This talk presents two methodologies that improve model-informed experimental design and model-based prediction.
The first is a multi-objective optimization algorithm that identifies combination treatment protocols maximising synergy from the perspective of both efficacy and potency (toxicity), while reconciling the sometimes contradictory assessments made by different synergy metrics. The second uses parameter identifiability to ask what minimal amount of experimental data must be collected, and when, to have confidence in a model's predictions. Real-world applications of both methodologies are presented.
Topics
- parameter identifiability
- experimental design
- cancer dynamics, broader topic: dynamics
- Mathematical Models of Cancer, broader topic: cancer
- treatment response
- model-informed experimental design, broader topic: experimental design
- model-based prediction, broader topic: prediction
- multi-objective optimization
- combination treatment protocols, broader topic: protocols
- toxicity