When I was in medical school, long before artificial intelligence (AI) was invented, I went through an extreme four-year period of factoid memorization. I learned. It seemed like a futile exercise at the time, but later in life this old knowledge was reconstructed each time I encountered a new disease. This form of human learning, using the brain as a computer, is the standard AI aspires to.
But now, artificial intelligence has progressed to the point where it becomes a very useful medical tool. In fact, a new AI tool known as ‘Sybil’ showed an early detection rate for lung cancer of 86-94%. It reportedly works by picking up irritations and small abnormalities in areas where lung cancer can develop, before these changes are visible to the radiologist. It relies on huge databases to know what’s wrong.
Concerns raised so far are the lack of diversity in the database and the possibility of overdiagnosis, noted by many experts, including Dr. Robert Cerfolio, chief of thoracic surgery at NYU Langone Health ( Full disclosure: I am a professor of medicine and medicine). The director of NYU Langone Health’s “Doctor Radio” said it showed “great potential.” Early diagnosis remains the key to saving lives from cancer, and AI can help with that.
Miriam Bredella, Ph.D., a distinguished professor of radiology at Harvard University, said last week in SiriusXM’s “Doctor Radio Reports” that AI’s key purpose in radiology is that thousands of studies (X-rays, CT scans, MRIs) that Done for one reason, and using AI algorithms to find something else, such as the amount of saturated fat in your bones, which can be correlated with other health issues such as insulin resistance, diabetes, and osteoporosis. Done.
There are currently hundreds of image-specific AI algorithms across the fields of radiology and cardiology. The Food and Drug Administration (FDA) has approved more than 520 applications from 2019 through January of this year.
But what about direct clinical applications? According to a recent study on the clinical use of AI in osteoporosis, published in Nature, “By applying AI algorithms in the clinical setting, primary care providers can Classifying and recommending appropriate exercise programs can improve treatment.” Osteoporosis is often overlooked, so this is one example that certainly helps.
But I’m afraid there isn’t enough concern about AI’s limitations.
OpenAI’s ChatGPT has been found to accurately answer most medical testing questions, but who knows how much practitioners (or their patients) can rely on these models for real-time information exchange. yeah. Trying to rely on her AI answers without consulting a doctor.
Remember, AI works by pattern recognition, comparing what it sees to hundreds of thousands of previous databases, and answering questions posed by referencing these same databases. However, AI always lacks the flexibility and deep insight of a well-trained doctor.
Japan has led AI in healthcare with shared applications and AI diagnostic impressions via the cloud. The Japanese consortium, which includes Hitachi and Microsoft Japan (Iguana interface engine), has been up and running for several months, targeting small hospitals and rural areas in Japan where there is a shortage of doctors, and AI is helping doctors make diagnoses. and leverage the cloud to make doctor diagnoses. data sharing. While this sounds ideal, an online survey in Japan published last year found that patients were more interested in their practices than their doctors and expressed concerns about regulation and lack of accountability.
AI in healthcare applications shows no signs of stopping. According to the World Economic Forum, AI applications in US clinical practice have tripled since 2020. Research and Markets estimates that the global market for AI healthcare solutions will exceed $200 billion by 2030.
Here in the US, AI-led seminars are being held at many universities and medical centers, including MIT and NYU Langone Health, as AI prepares for clinical use as healthcare leaders follow Japan. . At MIT, a seminar scheduled in May will not only provide a foundation of understanding, but also teach the healthcare professional how to develop his AI product to meet his needs. However, Nader Mherabi, chief digital and information officer at NYU Langone Health, said generative artificial intelligence applications “have the potential to transform surgery, patient care, and education and research missions, but there are still many more ways to go before they can be used.” It requires development and careful consideration,” warned faculty.
The world would be wise to follow Merabi’s warning directives.
Marc Siegel MD is Professor of Medicine at NYU Langone Health and Medical Director of Doctor Radio. He’s his Fox News medical correspondent and author of a new book. The politics of fear and the power of science. ”
Copyright 2023 Nexstar Media Inc. All rights reserved. This material may not be published, broadcast, rewritten or redistributed.
