It is strictly regulated Healthcare spaces have historically slow adoption of new technologies, but artificial intelligence has ultimately invaded the clinical environment. This technique is now useful for determining whether colonoscopy results are malignant or benign, addressing infertility issues, transcription of doctor notes on the spot, for example.
“Early every year, AI tools are becoming more and more commonly used,” says Regina Barzilay, a well-known professor of AI and Health at MIT School of Engineering.
Barzilay and Harvard Law Professor I. Glenn Cohen are part of a team teaching the upcoming MIT Sloan executive education course, “Transforming Healthcare with AI.”
In a recent webinar, Barzilay said it's an exciting time to take your medication. “Today, we know that AI tools can make predictions that the human mind cannot make. Sometimes machines can look at images and genetic data to make predictions that a doctor or human expert cannot make,” she said.
As AI continues to permeate healthcare, hospitals and health professionals need to make technical decisions that will impact the organization's future. The amount of automation you implement depends on your organization's needs and goals [of] How do we use AI?” said Barzilay.
“Unfortunately, we're not yet in terms of blindly applying AI across the pipeline,” she said. “One of the goals of the course is to look at different scenarios and see all the different ways AI can affect patient outcomes, safety, efficacy and costs. With caution, we can combine the technologies currently available with the issues.”
Below are three ways that AI can help empower clinicians today:
1. Read images more accurately than humans
AI can be used to give humans clear confidence rather than eliminate unnecessary treatments. It also allows you to raise the alarm early when further treatment is needed.
For example, AI can “very accurately” identify patients who are likely to develop breast or lung cancer in the near future, Barzilay said. That's because AI often does a better job of discovering disease activity through scans. People can only find anomalies that are large enough to the human eye. When disease activity is less noticeable, “AI is clearly doing better than humans,” Barzilay said.
Previous detection of illness gives practitioners a wider choice in treating patients. Clinicians can also identify small subsets of the population that require follow-up care.
2. Eliminate the hassle of taking notes
Doctors who use AI to take notes can focus more on patient care.
“I know it takes a lot of time for clinicians to do the document. They don't like it,” Baltzilai said. “The important literature shows that it contributes to clinician burnout.”
She said that natural language and speech recognition technologies are well improved and that she has the ability to transfer notes almost completely. From there, she said all doctors have to do is review and sign off the transcript “that takes much longer than writing and recording everything.”
“Even if there are some mistakes, clinicians can tolerate them because they need to review and sign them,” Barzilay said.
3. Provide guidance tailored to individual patients
Barzilay said that AI can employ patient history and biological data and integrate information to predict the side effects of taking the drug. AI is good at it, suggesting, “For this patient, we really need to be careful about these side effects,” she said.
Barzilay compared the process to online shopping. There, companies can “provide me something very specific based on what they have bought in the past, what they have not bought in the past, what they have seen, etc.”
Additionally, Barzilay said that AI can help predict disease trajectories in individual patients. It “can very well recommend us what will happen and how your trajectory may evolve,” she said.
AI is also used in non-clinical settings. Oncology researchers, for example, use molecular modeling to better understand disease mutations and design more effective drug-based treatments. “That's where AI is now leading the way in science frontiers,” Barzilay said.
However, she warned that doctors should remain skeptical of AI, not relying on it, and not “degenerate their capabilities” to make healthy clinical decisions.
“You still need to keep your mind as a human and know what to see,” Baltzilai said.
Executive Education: AI-driven healthcare transformation
