MLATE: Machine learning for predicting cell behavior on cardiac tissue engineering scaffolds.

Area of Science:

• Biomaterials Science
• Regenerative Medicine
• Computational Biology

Background:

• Cardiovascular disease (CVD) is a leading global cause of mortality.
• Cardiac tissue engineering (CTE) offers a promising regenerative approach for CVD.
• Current CTE research faces challenges due to costly and time-consuming experimental validation of scaffolds.

Purpose of the Study:

• To develop and validate machine learning (ML) models for predicting cell behavior on cardiac tissue engineering scaffolds.
• To create a comprehensive dataset of CTE scaffold specifications from existing literature.
• To identify the most effective ML algorithms for optimizing scaffold design and performance.

Main Methods:

• A novel dataset of CTE scaffold specifications (materials, cell lines, fabrication methods) was compiled from literature.
• Scaffold performance was rated based on cell viability, proliferation, and differentiation (0-3 scale).
• Twenty-eight ML algorithms were evaluated, with ensemble methods further optimized.

Main Results:

• The XGBoost algorithm achieved 87% accuracy in predicting cell behavior on CTE scaffolds.
• Ensemble learning, specifically AdaBoost and Voting Classifiers, improved prediction accuracy to 93%.
• An open-source AI software, MLATE, was developed and published with a user guide.

Conclusions:

• Machine learning, particularly ensemble methods, can accurately predict cell behavior on CTE scaffolds.
• The MLATE software provides a valuable, cost-effective tool for accelerating CTE research and development.
• This data-driven approach has the potential to significantly advance cardiac tissue regeneration strategies.