Researchers at NYU Langone Health have demonstrated that artificial intelligence can generate explanations of heart test results that are accurate, relevant, and easy for patients to understand.
With new laws mandating immediate access for patients to their test results, AI-based cardiology reports explaining these results promise to be a great solution for reducing patient anxiety and lightening the workload for doctors and other medical personnel.
Image Credit: fizkes via Shutterstock / HDR tune by Universal-Sci
Echocardiograms, which use sound waves to create images of the heart's blood flow, are essential for diagnosing heart conditions. However, the complex and technical nature of these reports often leaves patients confused and worried. The NYU study (published in the peer-reviewed journal JACC: Cardiovascular Imaging) aimed to address this issue by testing an AI tool's ability to produce patient-friendly explanations of these test results.
Accuracy of AI-generated health reports
Using OpenAI's latest generative AI tool, GPT-4, the research team conducted their study with one hundred doctor-written echocardiogram reports. AI-generated reports explaining the echocardiogram results were evaluated by five board-certified echocardiographers. Impressively all AI explanations were rated either “all true” (84%) or mostly correct (16%).
In terms of relevance, 76% of explanations were judged to contain “all of the important information,” 15% “most of it,” 7% “about half,” and 2% “less than half.” None of the explanations with missing information were rated as “potentially dangerous’’.
Dr Lior Jankelson, associate professor of medicine at NYU Grossman School of Medicine (one of the authors of the study), stated that their study shows that generative AI models can be effective in helping clinicians explain echocardiogram results to patients. "Fast, accurate explanations may lessen patient worry and reduce the sometimes overwhelming volume of patient messages to clinicians."
Mandated instant access to test results for patients
The developments in this area are particularly timely in the United States due to the federal mandate from the so-called 21st Century Cures Act, which requires the immediate release of test results to patients.
While this law enhances patient access to their health information, it has also led to increased anxiety as patients struggle to understand their results without immediate guidance from their doctors.
The study highlights that AI-generated explanations can fill this gap, providing patients with instant clarity and more peace of mind. In one part of the study, participants without clinical backgrounds found 97% of the AI-generated rewrites easier to understand than the original reports, significantly reducing their worry.
The importance of human oversight
However, as stated above, the research also found that 16% of the AI explanations contained inaccuracies, underscoring the importance of human oversight. In one instance, the AI mistakenly described a small fluid accumulation around the lung, an error known as an AI "hallucination." Thus, while AI tools are promising, they are meant to assist rather than replace human clinicians.
Looking Ahead: The Future of AI in Healthcare
Dr. Jacob martin, the study’s first author, stated that their next step will be to measure the impact of explanations drafted by AI and refined by clinicians on patient anxiety, satisfaction, and clinician workload. The team plans to integrate these AI tools into clinical practice, aiming to enhance patient care and reduce the burden on healthcare providers.
As AI continues to evolve, its role in healthcare looks set to expand, offering new ways to improve patient experiences and outcomes. If you are interested in learning more details about the study, be sure to check out the paper listed below.
Sources and further reading:
Evaluating Patient-Oriented Echocardiogram Reports Augmented by Artificial Intelligence (JACC Cardiovascular Imaging)
Association of Immediate Release of Test Results to Patients With Implications for Clinical Workflow (National Library of Medicine)
Too busy to follow science news during the week? - Consider subscribing to our (free) newsletter - (Universal-Sci Weekly) - and get the 5 most interesting science articles of the week in your inbox
FEATURED ARTICLES:
