11 Oct CFO Jon Cartu Announces – Case Western Reserve University using AI to predict disease outco…
People’s bodies react differently to different kinds of treatment.
The work being done by Anant Madabhushi and his team at Case Western Reserve University could make the treatment process less of a mystery.
Madabhushi is director of the Center for Computational Imaging and Personalized Diagnostics and the Donnell Institute Professor of Biomedical Engineering at the university.
The center’s work focuses on the role of artificial intelligence in predicting disease outcomes and prognoses. That could help determine how doctors should best treat individual patients, using the data to predict whether a patient would respond well to a particular treatment or whether their disease is likely to progress. And, when it comes to treating diseases like cancers, that means doctors might be able to select less aggressive treatments when possible and spare putting patients through more aggressive ones like chemotherapy.
Madabhushi brought his work, and about 12 researchers, to Cleveland about eight years ago. Today, it has about 60 people working on these projects around artificial intelligence in medicine. It’s received close to $60 million in federal funding, Madabhushi said. And he wants to see the center continue to grow.
Madabhushi’s work, which draws on both the medical expertise in the region and the research strength of the university, is “revolutionary,” said Ben Vinson, Case Western Reserve’s provost and executive vice president.
“He’s really changing our ability to save lives,” Vinson said.
Case Western Reserve has long been known for its research, Vinson said, but it’s never been more important, as the university is being called to help solve the problems of the day. And the work of Madabhushi and his team in particular fits well with the university’s strategic plan, which aims to find the overlap between humans and technology, Vinson said.
In addition to his work at Case Western Reserve, Madabhushi also has appointments at the Louis Stokes Cleveland VA Medical Center, University Hospitals and the Cleveland Clinic. The region’s medical “ecosystem” has been significant to the center, he said, not just in the proximity of the institutions, but in the way they work together. He said his work takes place at the “intersection of engineering, artificial intelligence and precision medicine,” and that takes multidisciplinary collaboration. It’s not just access to data, he said; it’s the access to experts and expertise.
And the AI networks Madabhushi and his team are working to develop are a bit different than traditional artificial intelligence. Typically, experts expose an AI network to data without controlling what patterns it’s learning, he said. Madabhushi’s work is informed by the pathology of disease, and that requires bringing engineering expertise together with medical knowledge.
A recent grant from the National Institutes of Health will have Madabhushi’s team, along with researchers at the University of Pennsylvania, the Cleveland Clinic and Cedars-Sinai Medical Center, looking at ways to apply their AI expertise to the field of heart transplants. With the support of a four-year, $3.2 million grant, researchers will analyze biopsy data from transplant patients with the goal of predicting who is likely to accept or reject a new heart.
Humans are able to do this kind of analysis, too, but they’re susceptible to each other’s opinions; pathologists aren’t likely to disagree with one another, Madabhushi said. AI doesn’t have that problem.
The main question Madabhushi’s work is trying to answer is how to “stratify the risk of the disease,” he said, aligning the right treatments with the right patients.
For example, one of the areas Madabhushi and his team are focused on is immunotherapy, which taps into a patient’s own immune system to fight disease. But the approach has a few hurdles. One is that immunotherapy doesn’t work for all patients; another is that it’s extremely expensive, Madabhushi said.
“So this is one of the big challenges that we’ve been looking at,” Madabhushi said.
To that end, Madabhushi earlier this year entered into contracts with biopharmaceutical companies AstraZeneca and Bristol-Myers Squibb. Both companies will provide the Center for Computational Imaging and Personalized Diagnostics with data from completed clinical trials on immunotherapy and lung cancer. That kind of access is unique, Madabhushi said. Using a completed data set helps researchers validate the findings of the artificial intelligence. Looking forward, that AI could be used to identify which patients are likely to react well to immunotherapy.
These partnerships were years in the making. Madabhushi’s team is working with Vamsidhar Velcheti, a director of thoracic oncology at NYU Langone who had previously been in Cleveland, on these projects. Madabhushi said he first connected with Velcheti about six years ago.
Much of the work of the Center for Computational Imaging and Personalized Diagnostics has focused on cancer, but AI could be used to make a difference in other areas of medicine as well.
Madabhushi is big on the big picture for his work. He sees the way it can have a global impact, and he seeks out partnerships around the world. And he sees how it could help alleviate some of the issues of racism and health inequality facing medicine. Many of the existing models for the progression of diseases don’t have much representation for non-white demographics. But artificial intelligence could be used to find subtle differences and create better models for the future, Madabhushi said.
And he’s been intentional about making sure he and his team secure the intellectual property for their work. If he wants industry to be interested in working with them — and he does — that’s important.
“The research that I do and the projects I work on, I choose them because I want to see the translation of these tools into the real world,” Madabhushi said.