Biomedical startups are using artificial intelligence to predict the response patients will have to cancer treatments, aiming to increase the success of drugs in clinical trials and tailor therapies to individuals. As data accumulate from clinical trials and fields such as gene and protein research, AI is helping scientists sift through large volumes of information to uncover signatures that correlate with response—or resistance—to treatment. Startups are using it to predict which drugs are likely to work in clinical studies and create tests to help doctors choose treatments.
AI allows researchers to pull diverse data sets together and avoid the biases that can stem from more limited data collections, said Terri Shieh-Newton, a member of law firm Mintz, Levin, Cohn, Ferris, Glovsky and Popeo, who advises life-sciences companies and holds a Ph.D. in cellular and molecular medicine. On its own, AI doesn’t assure success.
Humans must make careful decisions about the data sets they use to train algorithms,Shieh-Newton said. “There’s a human judgment that’s really needed," she added. “With machine learning, it’s easy to get off on the wrong foot." Venture firms are banking on the expertise of founders—and their ability to get access to the data needed to develop highly predictive tests.
Read more on livemint.com