By implementing AI in radiology, physicians can advance the medical field
- AI is at the forefront of radiology, but clinicians are still adapting to its use.
- Technological advances may one day help predict response to disease.
- Proper training can prepare future physicians for this rapidly changing field.
- This article is part of the “Innovation in Healthcare” series, highlighting what healthcare professionals need to do to cope with this technological moment.
Developments in artificial intelligence are opening up new possibilities in radiology. Advances help doctors make more accurate diagnoses and predict patient outcomes.
“It’s a really exciting time for medicine,” said Dr. Yvonne W. Lui, associate professor of radiology and associate chair for artificial intelligence at NYU Grossman School of Medicine. “We are seeing the implementation of artificial intelligence or machine learning tools across disciplines, and diagnostic imaging has always been at the forefront of these technological advances.”
In a 2020 member survey, the American College of Radiology (ACR) found that 30% of radiologists use AI in clinical practice to improve image interpretation of intracranial bleeding, blockage of arteries in lungs and breast abnormalities. The 20% of practices not using AI said they plan to integrate the technology into the clinic within the next few years. ACR’s Data Science Institute also publishes a running list of FDA-approved AI radiology technologies.
Radiology is still adapting to AI tools in the clinic, Lui said. “There’s no roadmap right now. I think people are trying to figure out what’s the best way forward and how to use certain tools.”
According to Dr. Anant Madabhushi, director of the Center for Computational Imaging and Personalized Diagnosis at the Case School of Engineering in Cleveland, diagnosing patients could be the beginning of the potential of AI in radiology. “The opportunities around treatment management, monitoring and response prediction are going to be potentially even more important than the diagnostic implications,” he said. Training the next generation of radiologists in AI is important to ensure the technology is properly applied in the clinic, he added.
Diagnose patients beyond the disease
Much of the buzz around AI in radiology has focused on classification tools. For example, clinicians use AI to determine whether a result on an image is normal or abnormal, such as whether there is pneumothorax – or air leaking between the lungs and chest wall – or no pneumothorax. .
This technology can help clinicians analyze millions of tiny pieces of information. Researchers are developing an AI that can speed up and make the task of interpreting an image more accurate. Quantification and identification of tumor volume, new lesions and tissue changes over time can be particularly helpful, she said.
Scientists are also trying to figure out how AI can evaluate images in a new way. “We ask if the images contain information that we have not previously been sensitive to,” Lui said. His own research is evaluating whether AI can help detect a signal in images of concussion patients to help identify injuries and monitor recovery.
AI may be able to help predict cellular response to cancer treatment and treatment monitoring, which could help avoid the need for invasive biopsies, Madabhushi said. “There are several different scenarios where the features of the imaging are so subtle that it really becomes impossible to distinguish between a treatment confounder and a disease,” he said.
For example, benign radiation necrosis, which is a lesion that forms at the site of the tumor due to radiation therapy or chemoradiotherapy, can be difficult to distinguish from tumor recurrence. AI can be an important supporting tool for radiologists to make this distinction and avoid the need for invasive biopsy, Madabhushi said.
In his own research, Madabhushi is investigating how AI in radiology can noninvasively identify prostate cancer and stratify the risk of its aggressiveness on an MRI to help guide treatment decisions. “What we’re trying to do is actually create a virtual biopsy,” he said.
AI in radiology training
The Radiologic Society of North America offers continuing medical education in artificial intelligence and imaging, webinars, and an AI certificate program. The American Academy of Radiology’s Data Science Institute (DSI ACR) also offers lectures on the use of AI in clinical practice, evaluation of AI algorithms, and a variety of online use cases. .
NYU Grossman School of Medicine offers a formal RN program in biomedical imaging for radiology residents. Courses in machine learning in radiology are also available for medical students and fellows, while research opportunities in radiology are open to doctoral students.
Madabhushi encourages medical students interested in AI in radiology to seek opportunities within their institutions to work with biomedical engineers and computer scientists specializing in bioinformatics.
“You want students to have an idea of what it takes to train AI algorithms in radiology, how to interact with AI and understand its limitations,” he said. “This collaboration breaks down some of the misconceptions about what AI can and cannot do.”