Communications is much more than just words. A classic UCLA study by (now) Professor Emeritus Albert Mehrabian argued only “7% of message relating to feelings and attitudes is in the words that are spoken.” The remaining 93% of the message is derived from facial expression (55%) and the way the words are spoken (38%), according to that study. Whether or not Mehrabian’s proportions are exact, the implications for healthcare are clear. An industry that often begins client interactions asking, “So how do you feel today?” benefits from looking beyond the words in client responses. Reading others’ emotions accurately is tough, even with people with whom one is intimately acquainted; just ask any couple. If it’s true that the words used have minimal relevance to emotions, digital health tech intervention that relies solely on patient verbal response is likely doomed to fall short.
We’ve written before about health tech designed to monitor emotions. At CES 2015, Sensura highlighted its API for tracking emotions based on heart rate and galvanic skin response. We also wrote about MIT Media Lab spinoff mPath‘s Moxo, a wearable that detects the emotional response, also based on GSR and conductance. MIT, long associated with artificial intelligence study and development, also figures in Affectiva, a company with several emotion-aware AI platforms based on facial imagery. Rana el Kaliouby, Affectiva’s CEO and Co-founder, was one of MIT Technology Review’s 35 Innovators Under 35 in 2012. Affectiva’s Emotion SDK for Face supports major mobile and desktop platforms, using optical sensors or standard webcams. The SDK employs computer vision, machine learning, and deep learning AI methodologies to train algorithms that classify emotions. Affectiva’s SDK accesses a huge emotion database, a repository of 5.7 million faces analyzed in 87 countries. The algorithm analyzes both facial and vocal expressions to identify human emotions.
Affectiva offers a variety of emotion recognition products. In addition to Emotion SDK and API – Face, the company’s products include Affdex for Market Research, Emotion as a Service, Emotion API – Speech, and In-Lab Biometric Solution, a research platform with sensors and emotion recognition technology in one location. Robots as companions make tech headlines, but advances in emotion-aware digital tech have exciting potentials for medicine and healthcare.