In developing medical artificial intelligence (AI) systems, data scientists and doctors are often on opposite sides of a frustrating information divide. A new partnership with the Mayo Clinic will bring together a team of doctors and couples, Zhao Chieju and Jiang Chiachun, at Stanford University’s Human-Centered AI Institute (HAI) to learn how to bridge the gap.
Improving treatment for migraine patients
At the Mayo Clinic in Rochester, Minnesota, Chia Chun Chan, MD, seeks treatment for patients who have suffered acute stroke and severe headaches that can incapacitate them for weeks or months. I am in charge of treating migraine patients who visit me. Or more. An expert in headache medicine and vascular neurology, Chen wants to not only improve this everyday reality for patients, but one day she will explore how AI can improve overall care through exciting new breakthroughs. I also want to explore how we can connect.
Next year, Chen will be a visiting fellow at Stanford HAI to study the intersection of AI and neurology. The appointment is part of a new joint program between HAI and the Mayo Clinic that will give Chen the opportunity to do what she has always loved: finding innovative solutions to difficult medical questions. Become.
Born in Taiwan, Chiang Kai-shek found his medical horizons broadened early in his career by both his neurosurgeon father and a series of influential leaders on two continents.
“My father always encouraged me to think about advances in medical research—discoveries—and not just the everyday realities of medicine,” she says. “Arthur Chiou, a biophotonics professor at the National Yangming University, opened a window for me to see different types of medical research. I met Professor Chien, who introduced me to the world of using mathematics and physics to solve medical research problems.It was inspiring.”
Chiang Kai-shek’s time at Stanford HAI is devoted to several research projects focused on her special interest in migraine and stroke. She is using AI to help doctors determine migraine prophylaxis medications that can help prevent debilitating attacks in chronic migraine sufferers (those who experience more than 15 of her migraine attacks per month). want to be.
“At the moment, there is no way to predict whether a patient will respond to drugs A, B, C, D, or E,” she says. “It’s a process of trial and error, starting with one drug and waiting three months to see if it works. If you have pain, this can be a long time and a long process. Altitude “We want to apply advanced machine learning technology to the Mayo Clinic’s established headache database to help predict appropriate matches and develop personalized treatment plans for migraine sufferers.”
Cheng is also a professor of medicine at Stanford University School of Medicine to help make better use of all patient electronic medical record data to understand which migraine patients are at risk of stroke and cardiovascular disease. We are also working with Nigam Shah, chief data scientist at Stanford Healthcare. disease.
“Clinician notes, consultations, admission notes, etc. are written in free text, but have been difficult for data scientists to access,” she says. “Dr. Shah’s lab has developed several advanced machine learning tools tailored to medical records that can extract and use this type of information. We hope to be able to apply it to two different health care systems to get robust research results.”
Jiang said AI is transforming the field of neurology through innovations such as telemedicine and AI-assisted CT analysis, both of which are important for stroke patients, especially in isolated areas where small hospitals do not have a full-time neurologist. Said to be beneficial. But to continue to introduce relevant and user-friendly technology into healthcare, we need better communication between those developing AI and those treating patients.
“I am very excited about leveraging AI in my field. That is why I am excited to come to HAI and help this process by facilitating more discussion and research between data scientists and clinicians. I wanted to contribute,” she says. “I see suffering patients in my clinic every day. It is important for us to continue imagining how we can develop AI tools to improve their quality of life.”
Developing better tools for echocardiography
Mayo Clinic cardiologist Chao Chieju Chao was a teenager in Taiwan when college entrance exams determined his path to medical school. But deep inside, the 18-year-old had secret ambitions.
“I was happy to go into medicine, but I always loved physics,” he says. “It was always in the back of my mind. I love the laws of physics and the beauty of formulas, and when I was a kid I dreamed of being the next Einstein or Feynman.” and then a postdoctoral fellow at the University of California, San Diego, where he led a research project to build a 3D dynamic heart model. He began to see intersections between physics and medicine, especially in the field of cardiology. “It was one of the most important periods of my life and one that greatly changed my career path,” he says.
Zhao will have the opportunity to delve deeper into his interdisciplinary interests as a visiting fellow next year at the Stanford Human-Centered Artificial Intelligence Laboratory. Expands current research on the potential use of AI technology to enhance echocardiography, which uses ultrasound to examine heart anatomy and dynamics, as part of a joint program between Stanford HAI and the Mayo Clinic intend to do something. Chao hopes that his dual background in medicine and technology will help bridge the knowledge gap that often exists between medical and technology professionals and accelerate the expansion of AI systems into clinical medicine, such as the Mayo Clinic. doing.
“The usual procedure so far has been for clinicians to come up with ideas and talk to AI engineers,” he says. “However, that process can take a long time because communication between the two disciplines is not easy. This program by Mayo and HAI will be the first time that clinicians will be in the AI field to develop these skills. You learn some and you get it back.I knew this was what I wanted to do.I consider myself a person who can speak both languages.”
At the Mayo Clinic, Zhao explained how AI models can be used to answer clinical questions, such as how to classify the grade of diastolic dysfunction that occurs when the heart’s damaged ventricles don’t completely fill with blood. I am researching if I can use it. (In fact, that work earned Chao a finalist for the “Young Investigator Award” at the 2022 American College of Cardiology.)
Stanford HAI will focus on two major research projects. The first incorporates the use of AI to give cardiologists a more realistic view of the heart during surgeries such as aortic and mitral valve replacements, which are often performed using catheters rather than open-chest surgery. It provides a view. Positioning of these catheters currently relies on 2D images provided by echocardiograms and X-rays. Chao said an AI model could compile her CT image of a patient’s heart and its information, giving doctors real-time 3D and hologram-based images of him to make the procedure smoother and more efficient. I envision that there will be
“Our goal is to improve success rates with shorter surgeries, minimize complications, and reduce radiation exposure,” Zhao said. “This could reduce the operating time from 5-6 hours for him to 2-3 hours for him, allowing for more efficient use of operating room resources.”
Chao then explores potential uses for AI visual language models to more efficiently generate time-consuming echocardiographic reports. Ideally, the model should have the ability to perform an initial interpretation of the study and then generate a preliminary report that can be reviewed by clinicians.
“This will be an advancement of current AI applications in the medical field,” said Zhao. “People have been trying to create a system that can report a single image, a chest x-ray. Echocardiography deals with video his clips that record the dynamics of the heart from different perspectives. It It’s an even more exciting challenge.”
Zhao says he is motivated both by his colleagues at the Mayo Clinic who are passionate about the research and by the wide range of academic and technical resources he finds at Stanford HAI.
“It’s great to work with cutting-edge AI scientists and the latest technologies,” he says. “Talking about my clinical ideas with my colleagues in this environment has been a very different and immersive learning experience for me, giving me insight into what people are doing here, and potential insights I never thought possible. These conversations are generating new ideas as we learn about different applications.It’s all really exciting.”
Stanford HAI’s mission is to advance AI research, education, policy, and practice to improve the human condition. learn more.
