Design and evaluation of a system for multi-modal automated pain recognition

A monitoring system - in the form of a demonstrator - is to be developed (in NIATS, UFU, Uberlandia, Brazil), which will guarantee a valid recognition of pain in patients in an acute care environment. Signals for the detection and quantification of pain can have mimic, gestural, (psycho-)physiological and paralinguistic character. Pain-relevant parameters are recorded using non-invasive sensor technology, which records data on the (physical) reactions of the person in pain. This can be realized by a camera technology that records facial expressions, gestures or posture, while audio sensors record linguistics or paralinguistics. Physiological information can be derived via corresponding biopotential sensors (e.g. electrodes).