WVU Researchers Develop AI Models to Transform Heart Disease Diagnosis

Researchers at West Virginia University (WVU) are advancing the use of artificial intelligence (AI) to enhance the diagnosis and prediction of heart disease among rural patients. This initiative addresses a significant gap in healthcare technology, as most existing AI models are primarily designed using data from urban populations, which can lead to inaccuracies in rural settings.

Addressing Health Disparities with AI

According to Prashnna Gyawali, an assistant professor in the Benjamin M. Statler College of Engineering and Mineral Resources, the development of AI models has become increasingly common globally. However, he points out a critical issue: the majority of the data utilized in these models originates from urban areas, which often have different demographic and biological characteristics compared to rural communities.

This lack of representation in healthcare data hampers the effectiveness of AI in rural health systems. To counter this, Gyawali and his team have initiated a project focused on training a new AI model using exclusively anonymized patient data gathered from various regions of West Virginia. “You have to ensure your algorithms have seen the populations where you want them applied,” Gyawali explained. He emphasized the importance of tailoring AI tools to the specific characteristics of rural populations to ensure accurate diagnoses.

Potential Benefits for Rural Healthcare

The research team is currently assessing the AI models’ ability to diagnose heart disease using historical data. Gyawali believes that effective AI applications could significantly alleviate the burden on healthcare professionals in rural areas, where resources are often limited. “Health care problems are growing, and we have manpower shortages,” he stated. He noted that patients in West Virginia frequently face long travel times to access proper healthcare services, sometimes requiring hours to receive initial diagnoses.

By integrating AI into rural clinics, Gyawali envisions a system where affordable scanning devices equipped with AI could facilitate early detection of conditions like heart disease. “If we have more clinics with inexpensive scanning devices with an AI system attached, we can have an early detection system flagging certain patients,” he added. This proactive approach could lead to timely interventions and improved health outcomes.

Despite the promising prospects, Gyawali cautioned that the AI models have only been tested on historical datasets and have not yet been applied in real-world clinical settings. There remains a critical need for ongoing refinement of the models to ensure their safety and reliability. “Whenever we talk about safety-critical applications like health care, we need to make sure they’re reliable,” he said.

The team is committed to enhancing the AI model to ensure it meets rigorous standards before any clinical trials commence. Gyawali noted that while no timeline has been established for these trials, the research is progressing. “We’re adding more layers to ensure the model is reliable,” he stated. The team is exploring ways to enhance performance and validate the algorithms further, including potential collaboration with clinics outside of West Virginia.

Ultimately, Gyawali envisions a future where policy-level interventions can expedite the adoption of these AI tools in healthcare settings, paving the way for innovative solutions to address rural health disparities. “That’s the road map toward adopting these tools in clinics,” he concluded, highlighting the importance of thorough research and collaboration in advancing healthcare technology.