Researchers at Michigan Medicine have harnessed the power of artificial intelligence (AI) to create an algorithm that accurately predicts death and complications following a crucial heart procedure known as percutaneous coronary intervention (PCI). This innovative tool has the potential to revolutionize the way clinicians determine treatment for individuals with blocked heart arteries.
Percutaneous coronary intervention, or PCI, is a minimally invasive procedure used to alleviate blocked arteries by inflating a balloon and, in some cases, placing a stent to enhance blood flow from the heart. While it carries fewer risks than open surgery, such as coronary artery bypass grafting, PCI is not without its potential complications, including bleeding and kidney injury.
The key challenge lies in assessing the individual risk associated with PCI. Each patient’s situation is unique, and patients and clinicians have often faced difficulties estimating this procedure’s potential harm. This is where the newly developed AI-driven algorithm comes into play.
Precise risk prediction for informed decision-making
Lead researcher David E. Hamilton, M.D., a cardiology-critical care fellow at Michigan Medicine, emphasizes the importance of precise risk prediction in the context of PCI. He notes that this tool can recognize a wide range of outcomes post-PCI, offering care providers and patients the critical information they need to make informed treatment decisions.
Traditionally, risk stratification tools for PCI have existed, but many of them fall short in accuracy and do not involve patients in the development process. The Michigan Medicine team sought to rectify this by collecting data from adult patients who underwent PCI between April 2018 and the end of 2021, utilizing the Blue Cross Blue Shield of Michigan Cardiovascular Consortium (BMC2) registry.
The researchers employed advanced machine learning software “XGBoost” to analyze more than 20 pre-procedural characteristics, including age, blood pressure, and total cholesterol. This comprehensive approach allowed the development of an AI-driven model that demonstrated remarkable accuracy in predicting critical outcomes such as death, major bleeding events, and the need for blood transfusion. Notably, this AI model outperformed its predecessors that relied on similar pre-procedural characteristics.
Patient-centered and individualized approach
What sets this AI algorithm apart is its patient-centered focus. It incorporates PCI Patient Advisory Council feedback, ensuring it caters to individual needs and concerns. The objective is to provide patients and healthcare providers with a tool that facilitates collaborative decision-making and supports patient education regarding the potential risks associated with PCI.
With the prevalence of smartphones and electronic medical records, there is immense potential for integrating this AI-driven risk prediction tool into electronic health systems. This would make it easily accessible at the bedside, enabling providers to quickly relay complex information while enhancing patients’ understanding of the risks associated with PCI.
Accessible AI technology
The innovative technology developed by the Michigan Medicine team has been transformed into a user-friendly computer and phone application, ensuring widespread access and usability. This democratization of AI tools holds the promise of improved patient outcomes and enhanced shared decision-making between patients and clinicians.
Acknowledging the support provided by Blue Cross, Blue Shield of Michigan, and Blue Care Network as part of the BCBSM Value Partnerships program is essential. While Blue Cross Blue Shield of Michigan and BMC2 collaborated closely on this research, the opinions expressed by the author do not necessarily reflect the views of BCBSM or its employees.
Source: https://www.cryptopolitan.com/ai-outcome-of-heart-artery-procedures/