Abstract 9174: Integration of Nonlinear Analysis and Artificial Intelligence for Human Ipsc-based Drug Cardiotoxicity Classification

Abstract

Introduction: Recent advances in human induced pluripotent stem cell ( hiPSC ) technology have granted novel tools to evaluate drug-induced cardiotoxicity on human cardiomyocytes. The contractile signals recorded from hiPSC-derived cardiomyocytes ( hiPSC-CMs ) are highly complex and dynamic with great degree of characteristics to various drug treatment. However, traditional linear methods often fail to capture the subtle variations in these signals generated by hiPSC-CMs. In this work, we integrated nonlinear analysis, dimensionality reduction techniques and machine learning algorithms for better classifying the contractile signals from hiPSC-CMs in response to different drug exposure. Methods: Beating hiPSC-CMs were treated with different drugs and analyzed based on contractile motions and calcium flux. Nonlinear analysis was used to compute the dimensional properties. All the contractile parameters are incorporated in different machine learning algorithms to detect the variations in contractile behaviors that reflect cardiotoxic effects. The visualization of drug treatment and dosages were accomplished utilizing dimensionality reduction techniques, t-distributed stochastic neighbor embedding (t-SNE). Results: Machine learning algorithms combined with nonlinear analysis proved far more sensitive to the changes in contractile functions than when using traditional linear analysis alone. Using Radom Forest algorithm, classification between drug-treatment and baseline control showed an accuracy higher than 96% with the inclusion of nonlinear parameters. Higher dimensional relations between variables and treatment groups were successfully visualized utilizing t-SNE to gain further insights of each drug. Conclusions: Integration of nonlinear analysis and artificial intelligence has proven to be a powerful tool for analyzing cardiotoxicity and classifying toxic compounds through their mechanistic action.