West Virginia University Develops AI to Enhance Heart Disease Diagnosis

Researchers at West Virginia University are developing an advanced artificial intelligence model aimed at improving the diagnosis and prediction of heart disease specifically in rural populations. This initiative addresses a significant gap in healthcare where most existing AI tools are predominantly tailored to urban settings.

According to Prashnna Gyawali, an assistant professor in the Lane Department of Computer Science and Electrical Engineering, the majority of AI models rely on data drawn from urban environments that often do not reflect the biological and socio-economic realities of rural patients. This bias can hinder AI’s efficacy in rural healthcare systems, where access and resources are often limited.

In response, Gyawali and his team are focusing on training a new AI model using exclusively rural patient data collected from various regions in West Virginia. He emphasized the importance of ensuring that AI algorithms are informed by the characteristics of the populations they are intended to serve. “You have to ensure your algorithms have seen the populations where you want them applied,” Gyawali stated.

The research team has gathered anonymous patient datasets and is testing various AI models to evaluate their ability to diagnose heart disease based on clinical test results. Gyawali believes that if developed correctly, AI can greatly benefit rural healthcare systems. It would not only alleviate the burden on healthcare professionals but also facilitate the early detection of conditions like heart disease, enabling timely intervention.

“Healthcare problems are growing, and we have manpower shortages,” Gyawali noted. He described the challenges faced by residents who may need to travel several hours for an initial diagnosis. He envisions a future where clinics equipped with affordable scanning devices and AI systems could provide early detection for patients, significantly improving health outcomes.

Despite the optimism surrounding the project, Gyawali acknowledged that the AI models have yet to be tested in real-world clinical environments. The current focus remains on refining the technology to ensure its reliability and safety before it can be applied to patient care. “Whenever we talk about safety-critical applications like healthcare, we need to make sure they’re reliable,” he stated, highlighting the importance of avoiding misdiagnoses that could lead to inappropriate treatments.

The team is committed to enhancing the AI model’s performance. They are exploring how to validate the algorithms further and are looking for clinics outside their immediate study to test the model’s applicability across different datasets. Gyawali expressed aspirations for the project to extend beyond West Virginia, examining how the AI performs in other states.

“Finally, ensuring policy-level interventions is crucial so we can start trials on these algorithms in real-world clinical settings,” Gyawali added. While there is no set timeline for when clinical trials may begin, the research continues to evolve, aiming for a future where AI plays a pivotal role in diagnosing heart disease in rural communities.