AI in Medicine

The UCLA Health System generates a tremendous amount of data that can and is being used to inform medical practice and address patient care challenges. Artificial Intelligence in Medicine at UCLA (AI in Medicine) leverages advances in machine learning and AI, providing the capability to analyze vast amounts of data, including: 

  • The electronic health records of more than two million patients per year

  • Waveforms (such as EKGs) from more than 30,000 annual hospital stays

  • More than 1,500 daily imaging studies

  • The genetic data of more than 25,000 patients and growing 

UCLA AI in Medicine has been developed to identify high-risk cases, improve the speed and accuracy of diagnosis, predict treatment outcomes and side effects, and reduce medical care costs. Progress in this field requires innovative research in both AI and medicine; the expert team involved in this interdisciplinary effort is led by Eran Halperin, Ph.D. and represents the UCLA Departments of Anesthesiology and Perioperative Medicine, Biological Chemistry, Computer Science, Human Genetics, Neurology, Obstetrics and Gynecology (OB/GYN), Ophthalmology, Pathology and Laboratory Medicine, and Statistics. 

AI in Medicine Pilot Project 

In collaboration with the Department of Anesthesiology and Perioperative Medicine, UCLA AI in Medicine successfully built an accurate preoperative predictor of in-hospital mortality and complications using electronic medical record data of more than 50,000 patients. In 2019, AI in Medicine started to deploy this predictive model using machine learning in UCLA clinical practices to identify patients who are scheduled for elective surgery but are at high risk for complications. 

Now, with the success of this pilot project, the AI in Medicine team is broadening its efforts and applying the same principles to other clinical challenges at UCLA, with additional pilot programs in OB/GYN, ophthalmology, neurology, and the emergency medicine. Each of these collaborative projects will be strategically initiated in clinical care once the models are completed and exhaustively tested. 

 

Subscription for the UCLA Community:
Computational Medicine organizes several annual seminar series. Sign up here.

Subscription for the Public:
Computational Medicine has many exciting and informative events and meetings happening year-round. If you would like to stay in touch and learn more about these activities, subscribe here.

If you have general questions about Computational Medicine please email compmed@ucla.edu