Ambient intelligence in healthcare consists of creating smart hospital rooms featuring AI systems that perform a range of tasks to improve outcomes, such as sensors and AI that can immediately alert clinicians and patients when they have not sanitized their hands prior to entering the hospital room, according to the Sept. 9 news release.
Arnold Milstein, director of Stanford’s Clinical Excellence Research Center; Fei-Fei Li, co-director of the Stanford Institute for Human-Centered AI; and Albert Haque, Stanford graduate student, co-authored a recent paper in Nature that examined ambient intelligence in smart hospitals and smart homes, where they can help with remote care of elderly patients.
Ambient intelligence is based largely on the combination of two tech trends: available infrared sensors that are inexpensive and can be built into high-risk caregiving environments and the rise of machine learning systems as a way to use sensors and train specialized AI applications in healthcare.
The first type is active infrared, such as invisible light beams used by TV remote controls; however, these new active infrared systems can use AI to create a sort of radar that maps 3D outlines of a person or object. These infrared depth sensors are already being used outside hospital rooms for situations like detecting whether a person washed their hands before entering the room. In a Stanford experiment, researchers placed a tablet near the door that shows a green screen that flicks to red or another alert color if someone walks through and hadn’t washed their hands prior. The team chose a visual alert rather than audible to meet clinicians’ preferences.
“Hospitals are already full of buzzes and beeps,” Dr. Milstein said in the news release. “Our human-centered design interviews with clinicians taught us that a visual cue would likely be more effective and less annoying.”
The second type of infrared tech is a passive detector, such as the types that allow night vision goggles to create thermal images from infrared rays generated by body heat. In the hospital, a thermal sensor above an ICU bed would allow the AI system to detect movements like twitching or writhing beneath the patient’s bed sheets and alert clinicians.
The researchers have avoided using high-definition video sensors, such as the ones in smartphones, because capturing video may unnecessarily impede on clinician and patients’ privacy. However, Dr. Milstein said that the “preliminary results we’re getting from hospitals and daily living spaces confirm that ambient sensing technologies can provide the data we need to curb medical errors.”
More articles on artificial intelligence:
The future of AI in healthcare, according to 6 hospital innovation execs
US Military Health System taps Google Cloud to prototype AI predictive cancer diagnostic tool
Massachusetts General Hospital develops AI model that predicts patient’s 12-year lung cancer risk
