Artificial intelligence (AI) as a term has become generalized to the point of being meaningless, even within the technology industry. But there are multiple types of AI, each producing their own particular outcomes. Below are some of the specific forms of AI, how and what they learn, and the best use cases for each in virtual care.
Computer vision A recent survey shows that 84% of healthcare leaders rate patient safety and quality care as the most critical issues to address in their organizations. One form of AI, called computer vision, can “see” and capture images in video to allow a machine to identify and predict risk-related behavior. This technology can be used to monitor the rooms of patients who are at risk of falling.
Computer vision is able to identify the patient in a room, whether the patient is in their bed, near the edge of the bed, trying to get out of the bed, moving around the room, or close to the boundary of what is acceptable and safe in that room. Similarly, caregivers can use computer vision to prevent elopement, which is when a person under care leaves a room or facility without authorization or assistance.
Another way computer vision can enhance patient safety is by detecting signs of suicide ideation. The algorithm would be trained to identify visual behaviors that suggest a person is preparing to hurt themselves.
Staffing shortages at hospitals makes it difficult for nurses to make their rounds. Computer vision can be used for virtual rounding so nurses are free to perform other duties. And this technology also can be used to streamline patient flow. It can detect whether a room has a patient in it or is empty, as well as whether it’s been cleaned and is ready for another occupant.
Ambient listening Think about Siri, Alexa, and other virtual assistants that use a microphone to detect sounds and voices. This same technology can be applied to a healthcare setting as a protection against workplace violence.
In a recent survey cited by the American Hospital Association, 44% of nurses reported being the victims of physical violence. The Nurse Staffing Think Tank seeks to decrease rates of physical violence against healthcare professionals through federal legislation and/or regulations from the Centers for Medicare and Medicaid Services (CMS) requiring healthcare facilities to track workplace violence and put in place measures to ensure the physical safety of employees.
One of the ways to track and prevent physical attacks on healthcare workers is by deploying ambient listening in patient rooms and common spaces. By analyzing and understanding the context of language – are specific words being used, are there changes in decibel level, etc. – ambient listening technology can help avert violence against healthcare facility employees by alerting security personnel that a staff member is at risk of being verbally or physically attacked.