Communications of the ACM
Does AI in Healthcare Need More Emotion?
Emotion AI or affective computing has not been widely applied in healthcare, as can be seen in a recent literature review.
Credit: Getty Images
Emotion artificial intelligence (AI) or affective computing applications in healthcare are rare, with many scientists skeptical of the technology's utility.
A team at Spain's University of Seville reviewed 156 papers that used AI in pregnancy studies, to explore development of a mobile application or wearable for detecting health conditions.
The researchers determined AI has been used for detecting and monitoring fetal health, congenital birth defects, preterm birth risk, and gestational diabetes over the last 12 years, albeit with little clinical testing.
The University of Seville team only found two AI studies including emotional parameters in medical literature: one on the relationship between pre-eclampsia and anxiety, and one about patient reactions to prenatal tests.
Ohio State University's Aleix Martinez said AI's generally poor ability to interpret emotions reflects the same shortcomings in humans, while the University of Seville's Andreea Oprescu sees value in including emotional parameters for many healthcare areas, like pregnancy.
From IEEE Spectrum
View Full Article
Abstracts Copyright © 2020 SmithBucklin, Washington, DC, USA
No entries found