Leveraging AI for the Benefit of Accurate, Efficient Health Technology Management

Posted June 9, 2019

To leverage data analytics, artificial intelligence (AI), machine learning, and predictive modeling to enhance equipment management activities, the field of healthcare technology management (HTM) first must cleanse and standardize device data, according to presenters at a session on Sunday at the AAMI Exchange in Cleveland, OH.

After Elma Kotapuri, a student in biomedical engineering at Wright State University, provided an overview of advances in predictive analytics, Rita D'Angelo, PhD, president and CEO of D’Angelo Advantage and an adjunct professor in the School of Professional Studies at Villanova University, explained how AI is being used in laboratory medicine. She described how the processing and analysis of white blood cells can help fight cancer in humans, providing an edge in terms of speed and accuracy.

Adapting AI within mammography screening is allowing clinicians to predict, with greater accuracy, malignant and benign tumors—without putting patients through painful biopsy procedures—said Eliezer Kotapuri, CCE, PEng, PhD(c), chief clinical technology officer at Mass Technologies. Early-warning predictive modeling systems also are being used within healthcare facilities to alert clinicians about possible declines in patients' conditions, well before their condition deteriorates, he added.

Kotapuri further emphasized that the first step in the road toward realizing the benefits of AI is cleansing one's data. From there, the data can be randomized, trained, cross validated, and tested.

Although health technology is highly connected, the actual utilization of technology is not fully understood, explained Kotapuri. Connectivity and integration, he said, are only used for sending electronic patient health information, whereas data on the use of use of devices are not being captured and used effectively.

This current reality is unfortunate, he added, because through the application of predictive analytics, hospitals could accurately determine their technology needs from a variety of perspectives. Gaining an accurate picture of equipment inventory, said Kotapuri, is the first step in realizing the benefits of AI, which could be a boon for HTM.

Further, AI could provide insightful data related to numerous measures, including equipment age, cost of acquisition, total cost of ownership, maintenance, labor and parts, service history, failure rate, and uptime and downtime, added Sean Connolly, radiology services manager for the University of Maryland Medical System.

Ultimately, said Kotapuri, possessing this information would be valuable in terms of replacement planning and anticipating future technology needs.

"There are lots of fish in the water, but with its current model, HTM is trying to just catch one fish at a time. AI applications will allow us to catch a lot more fish in one go," said Kotapuri.