Pay attention to the media coverage around artificial intelligence, and it’s easy to get the sense that technologies such as chatbots pose an “existential crisis” to everything from the economy to democracy.
These threats are real, and proactive regulation is crucial. But it’s also important to highlight AI’s many positive applications, especially in health care.
Consider the Mayo Clinic, the largest integrated, non-profit medical practice in the world, which has created more than 160 AI algorithms in cardiology, neurology, radiology, and other specialities. Forty of those have already been deployed in patient care.
To better understand how AI is used in medicine, I spoke with John Halamka, president of Mayo Clinic Platform and a physician trained in medical informatics. Halamka explained, “AI is just the simulation of human intelligence via machines.”
Halamka distinguished between predictive and generative AI. The former involves mathematical models that use patterns from the past to predict the future, while the latter uses text or images to generate humanlike interaction. Predictive AI is most valuable in medicine today.
Predictive AI can analyse the experiences of millions of patients to help answer the question: “What can we do to ensure that you have the best journey possible with the fewest potholes along the way?”
For example, if someone is diagnosed with Type 2 diabetes, an algorithm can predict the best care plan for that patient based on various factors such as age, geography, medical conditions, and nutritional habits.
The Mayo Clinic team has partnered with clinical systems across the United States and globally, including in Canada, Brazil, and Israel, to improve the quality of their algorithms. By the end of 2023, Halamka expects the network to include more than 100 million patients’ medical records (with identifying information removed) to enhance care for others.
Predictive AI benefits
Predictive AI also enhances diagnoses. For instance, in colon cancer screening, gastroenterologists traditionally perform colonoscopies and manually identify precancerous polyps. However, studies estimate that 1 in 4 cancerous lesions are missed during screenings.
Predictive AI can significantly improve detection by identifying polyps during the colonoscopy and alerting physicians to take a closer look. In a trial across multiple centers, using AI reduced the miss rate of potentially cancerous lesions by over half.
Halamka believes that within the next five years, not using AI in colorectal cancer screening could be considered malpractice. He emphasises that AI doesn’t replace doctors but provides additional insight, enabling them to see more patients and across more geographies.
Generative AI, on the other hand, is a different story. Halamka highlights concerns over its quality, accuracy, and potential for producing inappropriate and misleading information due to the un-curated materials used for training. While generative AI holds promise in reducing administrative burden, its use in clinical care requires further validation and improvement.
There needs to be federal oversight, similar to the Food and Drug Administration’s role in vetting medications, to independently validate algorithms and share results publicly. There must be efforts to prevent the perpetuation of biases in AI applications in health care.
AI’s use in medicine requires a cautious and thoughtful approach. It must be rigorously studied and carefully deployed, prioritising the principle of “first, do no harm.” Nevertheless, AI has incredible potential to enhance health care by making it safer, more accessible, and more equitable. — Washington Post
Leana S. Wen is a professor at George Washington University’s Milken Institute School of Public Health and author of the book “Lifelines: A Doctor’s Journey in the Fight for Public Health.”