Researchers at Stevens Institute of Technology have developed two AI-driven smartphone applications to detect signs of depression non-invasively. PupilSense uses eye measurements to identify potential depressive episodes with 76% accuracy, while FacePsy analyzes facial expressions and head movements to detect mood changes.
Depression affects nearly 300 million people worldwide, but detecting it can be challenging, especially when those affected don’t report negative feelings to friends or clinicians. To address this issue, Stevens professor Sang Won Bae is working on several AI-powered smartphone applications that could non-invasively warn users and others about potential depressive episodes.
PupilSense works by constantly taking snapshots of a user’s pupils, analyzing 10-second “burst” photo streams captured while opening phones or accessing social media apps. The system accurately calculates pupils’ diameters from the surrounding irises and compares them to determine abnormal responses. In one early test with 25 volunteers over four weeks, PupilSense proved 76% accurate at flagging times when people felt depressed.
FacePsy, another tool developed by Bae, analyzes facial expressions and head movements to detect mood changes. The system runs in the background of a phone, taking facial snapshots whenever it’s opened or commonly used applications are accessed. FacePsy has identified increased smiling as potentially linked to depression, which could be a coping mechanism or an artifact of the study.
The FacePsy pilot study found that fewer facial movements during morning hours and specific eye- and head-movement patterns were associated with potential depression. The study also revealed outward expressions of alertness or happiness can sometimes mask depressive feelings beneath. Bae concludes that FacePsy is a great first step toward developing a compact, inexpensive, and easy-to-use diagnostic tool.
These AI-powered smartphone applications offer a privacy-protective, accessible way to identify depression early, leveraging everyday smartphone use. The researchers aim to continue developing these technologies to help those affected by depression.
Source: https://neurosciencenews.com/ai-depression-facial-cue-eyes-27735/