At the UH Cancer Center’s AI Precision Health Institute, researchers have been using AI to improve cancer treatments and diagnoses since 2018. AI tools can search through records to find text-based medical clues that can lead to diagnosis, says John Shepherd, the institute’s co-founder and principal investigator and a professor at the UH Medical School. It can also help with second opinions on medical cases and lead local scientists to case studies. “We have a fairly low enrollment rate of cancer patients into clinical trials,” Shepherd says. “AI can search all cases in all medical centers in Hawai‘i for specific criteria for enrollment for hundreds of studies simultaneously.”
An example of his team’s research is the Hawai‘i Pacific Islands Mammography Registry, which uses AI to identify breast cancer trends, risks and outcomes in outcomes in Asian American, Native Hawaiian and Pacific Islander groups. The registry contains more than 45 million individual images—mammograms, ultrasounds and other procedures performed on 100,000 women over 10 years—connected to breast cancer outcomes and risk factors.