The Daily Pennsylvanian is a student-run nonprofit.

Please support us by disabling your ad blocker on our site.


The PRECISE Center, located within the School of Engineering and Applied Sciences, is researching possible uses of AI in the medical field. 

Credit: Sean Fang

The Penn Research in Embedded Computing and Integrated Systems Engineering Center is studying the use of artificial intelligence in medical and clinical practice. 

The PRECISE Center’s projects include enhancing vision assessments for at-risk groups and researching machine learning techniques for vision evaluation. The PRECISE Center is also employing multimodal AI to identify ocular abnormalities and diseases and establish reliable AI by designing systems that more consistently provide precise and reliable outputs.

Liz Wai-Ping Ng, the associate director of the PRECISE Center, told The Daily Pennsylvanian that PRECISE is undergoing industry-leading research pertaining to the role of AI in the medical field. This includes developing technology to expand healthcare access for traditionally underserved populations and gauging patients’ health wirelessly, according to Ng.

Ng also said that PRECISE sees three main implications of their work on generative artificial intelligence in clinical practice: data analysis, enhanced interpretability, and expanded accessibility. 

“Without abundant and diverse data, medical models might be plagued with poor performance or bias,” Ng said. “However, generative AI is promising in synthesizing realistic and diverse data, thereby extending the range of pre-existing datasets and enabling medical models to obtain superior performance.”

Ng added that generative AI provides a new way to improve accessibility for underserved populations. For example, it can increase access to glaucoma detection, particularly for Black communities.

 The AI can also predict visual field loss and progression to improve patient understanding.

Generative AI, particularly large language models, has the ability to interpret results using natural language and therefore enable clinicians to spend more time on their patients, Ng said.

Ng said that a potential issue with the Center's AI-based research is its implementation. She said there are some concerns about privacy and re-training but added that these concerns can be effectively overcome. 

“The foremost thing is to understand that AI systems will not replace professionals but rather a decision support tool to enhance patient care,” Ng said. “That being said, it would be beneficial to educate professionals in regard to the strength and limitations of current AI techniques to help them understand how it can aid them in patient care.”