On Thursday, Nov. 4, the University of Maine Artificial Intelligence Series hosted a webinar on the uses of artificial intelligence (AI) in the healthcare industry. The webinar was held over Zoom and featured guest speaker Dr. Behrouz Shabestari, the acting director of the Division of Health Informatics Technologies and the director of the National Institute of Biomedical Imaging and Bioengineering (NIBIB) National Technology Centers Program.
The UMaine Artificial Intelligence Webinar Series is a series of free webinars for the Maine business community, policymakers, attorneys, healthcare providers and other members of the public. Experts from various companies come to speak on a variety of diverse topics in this field.
“It is my great pleasure to introduce speakers today that are going to talk about artificial intelligence in healthcare,” the Associate Vice President for Research at UMaine Ali Abedia said.
The Director of the Center for Computational Pharmacology and Computational Bioscience Program Larry Hunter was supposed to present during the lecture but due to technical difficulties was unable to.
“I’m going to talk today about the applications of artificial intelligence at NIBIB,” Dr. Shabestari said.
According to Shabestari, NIBIB has four scientific program areas. These include the Division of Applied Science and Technology (DAST), the Division of Health Informatics Technologies (DHIT), the Division of Discovery Science and Technology (DDST) and the Division of Interdisciplinary Training (DIDIT).
Shabestari is responsible for running the Division of Health Informatics Technologies. This division looks at the study of AI through image processing, visual perception and display and point of care technologies and diagnostics.
“AI can change how we gather data, not just how we interpret the data,” Shabestari said.
According to Shabestair, there are applications of AI that can provide medical results 10 times faster.
“If we have an application that uses the AI tool to get the information and interpret the data, then we also have the network for getting information from the standard 2-d images, so why can’t we go directly from the data to the information,” Shabestari said. According to Shabestari, researchers at NYU Langone Health are currently trying to solve this answer as well and it is an area that he is very interested in.
Another application of AI that is becoming impactful in the healthcare industry is AI with CT Denoising. CT scans involve the use of radiation to capture images and according to Shabestari, the impact that AI is having on this field is huge. According to Shabestari, AI machine learning for CT scanning can bring the dosage of radiation down 75% for the patients. Shabestari noted that radiation is not good for the patients so being able to reduce it while still producing strong images is a huge advantage.
Shabestari also notes that a major advantage of incorporating AI is you don’t need to be an expert on AI to be able to use its applications.
“It is a major advantage of AI training to be able to do things without needing complete expertise,” Shabestari said. Shabestari noted that healthcare professionals can still gather results and information with AI applications without being an expert on it.
Shabestari also noted how AI applications are being incorporated into the fight against COVID-19.
“It was a very challenging application, but wanting good COVID[-19] tests has expedited the AI application because the FDA approved it faster and there was a lot of money put into it to get faster results,” Shabestari said.
Shabestari talked about an ongoing clinical trial that is in the works in which AI applications are being utilized in the assessment and management of COVID-19. It involves a device that patients blow into and then they put the results under a microscope and compare them to other viruses. The clinical trial is currently using the machine learning application of AI to identify and learn more about these viruses, and according to Shabestari, it has been very successful.
“I hope this will be able to get to the market soon,” Shabestari said.
Shabestari also noted that a major challenge that healthcare professionals are facing with COVID-19 is monitoring patients throughout the recovery period.
“If people have COVID[-19] and then they recover and then it attacks a different area of the body such as [the] liver or kidney you cannot continuously get a CT scan,” Shabestari said. He says an alternative that can be utilized when monitoring patients recovery is ultrasounds.
According to Shabestari, the benefits of utilizing ultrasounds is that it is beneficial for assisting less experienced physicians with identifying features for COVID-19 detection and diagnosis and provides 86% accuracy.
There was a brief question and answer period toward the end of the webinar. A question raised by a spectator of the webinar asked what skill sets are important for students that are interested in entering this field aside from technical expertise in AI.
“A program is anything you can learn but it is the understanding. I’m not an AI expert. I cannot write one line of code and understand it or the layer of the network, but I can clearly understand all of the applications,” Shabestari said. “AI is just an optimistic algorithm, so you have to learn how to apply the AI.”
For more information and to see more webinars from this series you can visit the department’s website at ai.umaine.edu/webinar.
