A new platform driven by artificial intelligence could enable patients to communicate with their medical devices in real time to improve their health, according to a study published Tuesday in the journal Nature Medicine.
Brian Litt, a professor of neurology, neurosurgery and bioengineering at the University of Pennsylvania, knew that neurological signals that can precede epileptic seizures could be detected by implanted devices — but not immediately communicated to the patient.
His team designed a system that incorporated the intuitive chat features of popular large language models, but within a secure, private and health-focused platform that allowed patients to receive personalized alerts and report symptoms in real time through an app on their phone. Patient feedback would continuously help train the system.
The study, conducted by researchers from Penn’s medical school and its Center for Neuroengineering and Therapeutics, focused on epilepsy, but the researchers said they believe the technology could be applied much more broadly, including for conditions that use insulin pumps, cardiac devices and treatments for neuropsychiatric disorders such as Parkinson’s disease, depression and addiction.
Litt gave the example of a defibrillator detecting changes that might indicate the threat of a heart attack when a person is shoveling snow: “Why couldn’t it tell you?”
“I think the strongest signal we got from patients,” said Zack Goldblum, a graduate student and co-author of the study, “was when they said, ‘Can I take this home?’”
The post With AI platform, patients could chat with medical devices to improve health appeared first on Washington Post.
