- 16th August 2018
- Posted by: Manolis
- Category: Blockchain
Organizations are already predicting life-threatening inpatient scenarios by leveraging machine learning. What else is on the horizon?
In a policy recommendation passed this year by the American Medical Association, the organization lauds the potential of artificial intelligence in healthcare. Combining AI with human clinicians can advance care delivery “in a way that outperforms what either can do alone,” the AMA says.
While such technology has been considerably hyped in recent years, the organization understands that with tempered expectations — and deployed in the right situations — AI can have a real impact on the industry.
AI Makes Radiology More Efficient
Many believe one of AI’s biggest impact areas will be radiology. A report published in August by Signify Research projects that the global market for machine learning in medical imaging will surpass $2 billion by 2023.
Notable strides have already been made on this front. Researchers at the Massachusetts General Hospital and Brigham and Women’s Hospital Center for Clinical Data Science (CCDS) in Boston, for instance, have trained a deep neural network that uses an integrated system designed for AI applications to sift through a library of 10 million radiology records. Experts expect the system will eventually be able to quantify biological tissue with more precision than a human.
At the University of California, Los Angeles, radiologists used deep learning to build a chat system dubbed “virtual interventional radiologist” that can answer many standard patient care questions for attending physicians and other clinicians.
Both tools hold promise for improving efficiency and assisting clinicians in providing better care, faster.
AI Improves Cancer Screening and Cardiac Arrest Prevention
Other healthcare areas are poised to take advantage of AI, as well. For instance, AI powers an innovative cancer-screening program currently in beta known as Doctor Hazel. The technology uses deep-learning neural networks to screen and classify skin cancers with 80 percent accuracy.
Additionally, earlier this year, Microsoft touted a new AI tool developed by electronic health record vendor Epic and powered by Azure to help clinicians at Ochsner Medical Center in New Orleans predict, and potentially prevent, patient coding — cardiac or respiratory-arrest situations. In a 90-day pilot last year, Ochsner reduced the typical number of codes by 44 percent.
What’s more, in a post to his cHealth Blog, Partners HealthCare Vice President of Connected Health Joseph Kvedar notes that AI likely will be critical to overhauling the current caregiving model. He’s also pointed out that the technology will help to enhance chronic care as it matures.
Today’s Artifical Intelligence Jumpstarts the Future of Care
A report published in December on behalf of the Department of Health and Human Services by JASON, an independent research group, stated that AI is beginning to play a role in transformative changes in healthcare, but also noted “significant challenges” to the use of the technology, including acceptance of AI in clinical practice.
Still, the AMA policy recommendation states that the organization will explore AI more closely, examining potential legal implications and encouraging education for physicians, medical students, healthcare professionals, administrators and patients.