Research on heart function and disease has historically been complex and time-consuming. To study disease and test new drugs more efficiently, a group of researchers at Columbia Engineering has developed BeatProfiler, a software that automates the analysis of heart cell function. By integrating various heart function indicators into one tool, such as contractility, calcium handling, and force output, researchers can quickly and accurately assess heart cell function and test drugs that affect heart function. The project leader, Gordana Vunjak-Novakovic, believes that BeatProfiler is a transformative tool that is fast, comprehensive, automated, and easily accessible to investigators and clinicians.
BeatProfiler is the first system to automate the analysis of heart cell function from video data, allowing researchers to distinguish between different diseases, assess the severity of diseases, and test drugs objectively. The software is open-source, allowing any lab to use it for free. This decision was made to disseminate the research results and gather feedback from academic, clinical, and commercial labs to further refine the software. The development of BeatProfiler was driven by a need to diagnose heart diseases more quickly and accurately, particularly as researchers were creating more advanced cardiac models to study diseases and potential therapeutics.
The team behind BeatProfiler collaborated with experts in software development, machine learning, signal processing, and user experience to create a graphical user interface that would make it easier for biomedical researchers to analyze data. The results of the study showed that BeatProfiler could analyze cardiomyocyte function accurately and quickly, outperforming existing tools by up to 50 times in some cases. The software was able to detect subtle changes in engineered heart tissue force response that other tools might miss and classify different cardiac drugs based on their effects on the heart with high accuracy.
Moving forward, the researchers are working to expand BeatProfiler’s capabilities for new applications in heart research, including a range of diseases that affect heart pumping and drug development. They aim to validate its performance across various in vitro cardiac models and refine the machine-learning algorithm to classify a variety of heart diseases and drug effects. Ultimately, the goal is to adapt BeatProfiler for use in pharmaceutical settings to expedite the testing of numerous candidate drugs simultaneously, potentially revolutionizing drug discovery and heart disease research.
