Computational tools have been applied to make the design of the drug and medical devices more rationale and personalized. Simulation models have been developed to reproduce the mechanistic details of cross-scale pathophysiology and predict patient response to treatment interventions. That said, simulation models are of limited practical use interpreting large data sets because high-fidelity simulations typically require a large number of parameters, some of which are unknown. In contrast, machine learning models are powerful to effectively process, integrate and analyse massive amounts of data. ML models are, however, limited to predicting the overall dynamics and identifying casual relationships of the biological system.
The goal of this Research Topic is to evaluate and discuss the potentials of integrating mechanistic simulation and machine learning in solving increasingly complex physiological and medical questions. As such, we aim to address the following questions:
- What are the challenges of existing modelling frameworks to answer complex, multi-scale physiological and medical questions?
- How can combinatorial computing/ modelling techniques improve the practice of personalized medicine?
- What are possible benefits and limitations of developing hybrid models in biology, bioengineering and biomedicine?
This Research Topic will feature a collection of articles that will be focused on:
• Integration of machine learning and mechanistic models in biomedicine
• Challenges of the hybrid model computing framework
• Multi-scale modelling in areas of physiology, bioengineering and medicine
• Machine learning as surrogate models of complex mechanistic models
• Parameter reduction and uncertainty quantification in machine learning and mechanistic models
Computational tools have been applied to make the design of the drug and medical devices more rationale and personalized. Simulation models have been developed to reproduce the mechanistic details of cross-scale pathophysiology and predict patient response to treatment interventions. That said, simulation models are of limited practical use interpreting large data sets because high-fidelity simulations typically require a large number of parameters, some of which are unknown. In contrast, machine learning models are powerful to effectively process, integrate and analyse massive amounts of data. ML models are, however, limited to predicting the overall dynamics and identifying casual relationships of the biological system.
The goal of this Research Topic is to evaluate and discuss the potentials of integrating mechanistic simulation and machine learning in solving increasingly complex physiological and medical questions. As such, we aim to address the following questions:
- What are the challenges of existing modelling frameworks to answer complex, multi-scale physiological and medical questions?
- How can combinatorial computing/ modelling techniques improve the practice of personalized medicine?
- What are possible benefits and limitations of developing hybrid models in biology, bioengineering and biomedicine?
This Research Topic will feature a collection of articles that will be focused on:
• Integration of machine learning and mechanistic models in biomedicine
• Challenges of the hybrid model computing framework
• Multi-scale modelling in areas of physiology, bioengineering and medicine
• Machine learning as surrogate models of complex mechanistic models
• Parameter reduction and uncertainty quantification in machine learning and mechanistic models
