West Virginia Researchers Harness AI to Transform Heart Disease Diagnosis

Researchers at West Virginia University (WVU) are developing artificial intelligence (AI) models aimed at enhancing the diagnosis and prediction of heart disease among rural populations. This initiative addresses a significant gap, as most existing AI health care models largely reflect data from urban areas, leading to potential biases that could hinder effective health care delivery in less populated regions.

Addressing Rural Health Disparities

According to Prashnna Gyawali, an assistant professor in the Lane Department of Computer Science and Electrical Engineering at WVU, the majority of AI models in use today rely on data sourced from urban settings. These urban datasets often originate from more affluent populations, which can differ biologically from those in rural areas. This bias can limit the effectiveness of AI applications in rural health care systems.

Recognizing this disparity, Gyawali and his team have embarked on a mission to train a new AI model specifically with data gathered from rural patients in West Virginia. “You have to ensure your algorithms have seen the populations where you want them applied,” Gyawali emphasized, indicating that tailored AI solutions are essential to accurately diagnose heart disease in rural communities.

Collecting Data for Effective AI Solutions

The team has amassed anonymous patient datasets from various regions within West Virginia. These datasets serve as a foundation for testing different AI models to evaluate their capabilities in diagnosing heart disease based on patient test results. Gyawali notes that if implemented effectively, AI could significantly alleviate the workload of overburdened medical professionals, helping to identify early warning signs of conditions such as heart disease, ultimately allowing for timely intervention.

“Health care problems are growing and we have manpower shortages,” he said. “In our state, we don’t have easily accessible health care infrastructures. A person who wants to get tested in a proper manner may have to travel several hours just for an initial diagnosis,” he added. Gyawali envisions a future where clinics equipped with affordable scanning devices and AI systems can facilitate early detection for patients, potentially transforming rural health care.

To achieve this vision, Gyawali stresses the importance of ensuring that AI models are both reliable and unbiased. “There are several checks we need before we reach that hypothetical — but beautiful — scenario for our state,” he remarked, highlighting the need for rigorous validation of the technology.

While the team remains optimistic about their current testing phase, Gyawali cautions that the AI models have thus far only interacted with historical datasets and have yet to be validated with real-world patient interactions. Ongoing refinement of these models is critical to ensure both medical and technical professionals can confidently utilize the technology in clinical settings.

“Whenever we talk about safety-critical applications like health care, we need to make sure they’re reliable,” Gyawali stated. “We don’t want to give medications to patients who are wrongly diagnosed. We want to make sure the model is identifying patients who require immediate care.”

Future Steps and Broader Implications

Gyawali and his team are committed to enhancing the AI model’s reliability before moving to clinical trials. Although there is no set timeline for this next phase, he assures that they are actively working on improving the model’s performance. “How can we further enhance performance? These are AI questions my lab is trying to answer,” he noted. The researchers are also exploring partnerships with clinics outside the study to test the algorithm on diverse datasets.

Moreover, Gyawali expressed a desire to expand the research beyond West Virginia, aiming to assess the model’s efficacy in other states. He advocates for policy-level interventions that would facilitate the adoption of these AI tools in clinical environments, emphasizing that such steps are essential for real-world application.

The work being done at WVU represents a promising advance in the integration of technology in health care, particularly for underserved rural populations. As researchers continue their efforts, the hope is to establish a framework where AI not only supports medical professionals but also significantly enhances patient outcomes in heart disease diagnosis and treatment.