SALT LAKE CITY — Dr. AI is starting a practice in a medical setting near you. And experts say that with this new attention to detail, your health care is likely to improve in a number of ways. But in some areas, you may want to take the clear route.
Artificial intelligence is fast becoming a staple in growing segments of healthcare, but it's not ready in others. Still, experts say you don't need to worry that you're missing out on the personal touch if you're getting it from a medical provider: as important as ever in the practice of human medicine. are You may even find that care providers have more time to attend to your needs.
The National Institutes of Health notes that artificial intelligence tools are revolutionizing a wide range of medical fields, including research, diagnosis and treatment. Convergent advances in computing power and the proliferation of large-scale health data sets are setting the stage for new approaches to research as scientists increasingly use AI software and tools to advance their work at a faster pace. using its powerful information processing capabilities.
“I'm excited about the technology,” said attorney Daniel J. Gilman, a senior scholar at the International Center for Law and Economics, a nonpartisan, nonprofit research center based in Portland, Oregon. “I think we've seen that as long as it's introduced and used in a careful and responsible way, AI seems to have tremendous promise.”
From experience to problem solver
Dr. Yves Lussier is both a doctor and an engineer—and an intrepid AI enthusiast. At the University of Utah School of Medicine, he is chairman of the department of biomedical informatics – the founding department of the field in the United States and perhaps the world, dating back to the late 1950s, he said.
Lussier traces AI's roots to the 1940s, when neural networks were developed to “cause uncertainty.” Later, AI moved towards reasoning with conviction. The pace of each AI advancement has been faster than the one before. By the mid-1970s, Stanford's software could reason with both certainty and uncertainty — “expert systems,” he said. Since, deep learning (15 years ago) and “transformers” (seven years ago) have given birth to emerging conversational AIs called “generative AI”, such as ChatGPT.
There are many types of AI. Most people don't realize voice recognition is the AI that unlocks your bank account. What Lussier called a “game changer” came seven years ago with Creative AI, which can be prompted to generate text, images, videos and other data. You can push a transcript into Creative AI and retrieve plain-spoken words or technical terms depending on your audience.
The biggest impact, perhaps, is helping to solve problems that have been largely intractable.
Dr. Nathan Blue is an obstetrician and assistant professor in the Department of Obstetrics and Gynecology at the University of Utah School of Medicine. Blue has been involved in research efforts for more than a decade, working to develop new clinical diagnostic strategies that can identify early signs of pregnancy complications that result from decreased placental content.
The deficiency can lead to complications including fetal growth restriction, bleeding, preeclampsia and stillbirth, Blue said. Traditional research strategies for reducing the risk associated with placental insufficiency have many pitfalls, he said, and can be crude and inflexible.
New research techniques that incorporate AI systems are showing promise for overcoming some of the shortcomings of previous strategies and could lead to new medical approaches that, Blue said, reduce stress for expectant mothers. Doing so will help reduce uncertainty about clinical intervention decisions and lead to better utilization. of resources.
“In the last couple of years, we've started working with the bioinformatics and genomics group here at the U.,” Blue said. “These senior thought leaders and investigators have helped us leverage more computationally advanced methods, including artificial intelligence, to better quantify risk.”
Part of the research work involves applying AI tools to large data troves, including anonymized genomic profiles of more than 10,000 obstetric patients, and zeroing in on diagnostic markers that can more accurately predict future high-risk pregnancies. I can be part of new clinical applications to help.
“In terms of research and investigation, what's really interesting about what AI-based tools and approaches can offer is that, until now, we've been trying different versions of the same thing,” Blue said. “Using fancy old-fashioned tools to find factors, but really we're walking around in the same sandbox, so to speak.
“What I'm most excited about in AI-based strategy is that it's helping us bypass many of the pitfalls in analytics but expanding how we use information. The applicability and accuracy of these tools is better than what we were getting before.”
Xiaodong Ma, a professor in the Department of Radiology and Imaging Sciences at the University of Utah School of Medicine, is a member of the Medical Imaging and Computational Analysis Lab, a research team working to develop advanced techniques that image AI is also involved in the acquisition, analysis and quantification of data. Research applications.
Among other projects, he and his team are investigating vascular problems — specifically how abnormalities in the carotid artery can serve as indicators of more serious vascular pathologies.
Analyzing images taken through MRI and/or CT scans has traditionally been a manual, time-consuming process, Ma said. Thanks to AI-powered, semi-automated image analysis techniques developed by the Medical Imaging and Computational Analysis Lab, the development of new diagnostic strategies has accelerated.
Another medical imaging project leveraging AI-based image analysis tools is looking at the relationship between calcification in the brain and the ravages of aging, he said.
“We have the ability to predict diseases associated with aging like vascular disease and Alzheimer's,” Ma said. “Our hope is that AI can help us screen these images and define which patients may be at higher risk.”
Workload triage
Women who have had Pap smears are already part of the story of AI in medicine. AI has been used for over 30 years to sort through millions of exams annually to determine which ones need special attention, proving its value there and in radiology, among others. AI can spot early signs of unhealthy tissue changes that might be missed by the naked eye, saving time, money and pain.
AI is designed to have many false positives and no false negatives. “It doesn't make a mistake in forgetting cancer. But it's 10 times more likely to claim cancer when there isn't one,” Lussier said. But how quickly AI moves through images makes cervical cancer screening manageable and affordable.
This type of artificial intelligence took a decade or more to design many years ago, Lussier said. Now, given the pace of progress, such a program could probably be built in weeks or months.
Will a redesign improve results? “No, it's highly accurate. But it will cost a lot less now because it will take less time and be done with better tools. It's a game changer,” Lussier said. Faster design using fewer resources means lower costs, more accessible to more users in industries like healthcare, providing greater consumer benefit.
Click here to read more about this article.