Can artificial intelligence overcome health system challenges? | MIT News

Even though rapid improvements in artificial intelligence have led to speculation of significant changes in the healthcare landscape, the adoption of AI in healthcare has been minimal. A 2020 survey by Brookings, for example, found that less than 1% of healthcare job openings required AI-related skills.

the Abdul Latif Jameel Clinic for Machine Learning in Health (Clinique Jameel), a research center within the MIT Schwarzman College of Computingrecently hosted the MITxMGB Conference on AI Treatments with the goal of accelerating the adoption of clinical AI tools by creating new opportunities for collaboration between researchers and physicians focused on improving care for diverse patient populations.

Once virtual, the AI ​​Cures conference returned to in-person attendance at MIT’s Samberg Conference Center on the morning of April 25, welcoming more than 300 attendees consisting primarily of researchers and physicians from MIT and Mass General Brigham (MGB).

MIT President L. Rafael Reif opened the event by welcoming attendees and talking about the “transformative capacity of artificial intelligence and its ability to detect, in a dark river of swirling data, the brilliant patterns of meaning that we could never see otherwise”. MGB President and CEO Anne Klibanski went on to praise the joint partnership between the two institutions and noting that the collaboration could “have a real impact on patients’ lives” and “help break down some of the barriers information sharing”.

Nationally, approximately $20 million of contract work currently takes place between MIT and MGB. MGB Academic Director and AI Cures Co-Chair Ravi Thadhani believes five times that amount would be needed to do more transformative work. “We could definitely do more,” Thadhani said. “The conference… only scratched the surface of a relationship between a leading university and a leading health care system.”

Regina Barzilay, an MIT professor and co-chair of AI Cures, echoed similar sentiments at the conference. “If we’re going to take 30 years to take all the algorithms and translate them into patient care, we’re going to lose patient lives,” she said. “I hope the main impact of this conference is finding a way to translate it into a clinical setting for the benefit of patients.”

This year’s event featured 25 speakers and two panels, most of them addressing barriers facing the mainstream deployment of AI in clinical settings, from fairness and clinical validation to regulatory hurdles. and translation issues using AI tools.

On the list of speakers, it is worth noting the appearance of Amir Khan, a principal investigator of the United States Food and Drug Administration (FDA), who answered a number of questions from curious researchers and clinicians. on the FDA’s ongoing efforts and challenges in regulating AI in healthcare.

The conference also covered many impressive advances made by AI in recent years: Lecia Sequist, an MGB lung cancer oncologist, spoke about her collaborative work with MGB radiologist Florian Fintelmann and Barzilay to develop a AI algorithm that could detect lung cancer up to six years in advance. MIT professor Dina Katabi showed MGB doctors Ipsit Vahia and Aleksandar Videnovic an AI device that could detect the presence of Parkinson’s simply by monitoring a person’s breathing patterns while they sleep. “It is an honor to collaborate with Professor Katabi,” Videnovic said during the presentation.

MIT Assistant Professor Marzyeh Ghassemi, whose presentation focused on designing machine learning processes for more equitable health systems, found the longer-term perspectives shared by speakers in the first panel on evolution compelling. of clinical science by AI.

“What I really liked about this panel was the focus on the relevance of technology and AI in clinical science,” says Ghassemi. “You heard some panel members [Eliezer Van Allen, Najat Khan, Isaac Kohane, Peter Szolovits] say they were the only person at a conference at their university that was focused on AI and ML [machine learning]and now we’re in a space where we have a miniature poster-only conference with people from MIT.

The 88 posters accepted for AI Cures were on display for attendees to peruse during the lunch break. The research presented covered different areas of interest, from clinical AI and AI for biology to AI-powered systems and others.

“I’ve been really impressed with the amount of work going on in this space,” said MIT professor Collin Stultz. Stultz also spoke at AI Cures, focusing primarily on interpretability and explainability risks when using AI tools in a clinical setting, using cardiovascular care as an example to show how algorithms could potentially mislead clinicians with serious consequences for patients.

“There is a growing number of failures in this space where companies or algorithms strive to be the most accurate, but do not consider how the clinician perceives the algorithm and their likelihood of failure. ‘use,” Stultz said. “It’s about what the patient deserves and how well the clinician is able to explain and justify their decision-making to the patient.”

Phil Sharp, a professor at the MIT Institute and chairman of the Jameel Clinic Advisory Board, found the conference energizing and thought the in-person interactions were crucial to gaining knowledge and motivation, unmatched by many conferences that are always hosted virtually.

“The broad participation of students, leaders and community members indicates that there is a realization that this is a tremendous opportunity and a huge need,” says Sharp. He pointed out that AI and machine learning are used to predict the structures of “almost everything,” from protein structures to drug efficacy. “He says to young people, be careful, there could be a machine revolution coming.”

Comments are closed.