Using AI to Improve Diagnosis and Treatment

Ahmad Tafti

The stresses on health care workers from COVID-19 compounded existing problems with medical errors. Now Pitt is at the center of a collaboration with Pittsburgh’s Jewish Healthcare Foundation to reduce medical errors by bringing together the medical, AI, and robotics expertise in Pittsburgh to develop autonomous patient safety technologies. Called the Regional Autonomous Patient Safety Initiative, the project aims not only to reduce medical errors but to put Pittsburgh on the map as a global hub in a new industry.

Pitt’s Department of Biomedical Informatics, and Schools of Medicine, Pharmacy, and Public Health are developing technology to reduce adverse drug events when senior patients transition between skilled nursing and other care facilities. At the core of the project is a decision support system bringing together multiple forms of data—such as medications recorded in a patient’s electronic health records—to train an autonomous AI-based technology in an easy-to-use format for providers.

“We hope to prevent medication errors such as accidentally overlooking someone’s depression medication in a care transition,” says Richard D. Boyce, associate professor of biomedical informatics and principal investigator on the project. “Transitions are an intervention point to create automated detection of medication errors.”

Elsewhere on campus, at Pitt’s HexAI Research Laboratory (where HexAI stands for Health + Explainable AI), led by Health Informatics Assistant Professor Ahmad Tafti, researchers and scientists from medical and rehabilitation disciplines are developing AI models to tackle clinical challenges. From total joint arthroplasty to stroke recovery, the groundbreaking research at the lab is improving the ability of physicians to provide more accurate diagnoses and effective treatments for their patients and helping rehabilitation providers assess the quality and consistency of intervention delivery in real-world settings.