Automatic Recognition Methods Supporting Pain Assessment: A Survey

Pain is a complex phenomenon, involving sensory and emotional experience, that often poorly understood, especially in infants, anesthetized patients, others who cannot speak. Technology supporting pain assessment has the potential to help reduce suffering; however, advances are needed before it can be adopted clinically. This survey paper assesses state of art provides guidance for researchers make such advances. First, we overview pain’s biological mechanisms, physiological behavioral responses, components, as well methods commonly used clinic. Next, discuss challenges hampering development validation recognition technology, existing datasets together with evaluation methods. We then present an all automated publications indexed Web Science from proceedings major conferences on biomedical informatics artificial intelligence, provide understanding current have been made. highlight progress both non-contact contact-based approaches, tools using face, voice, physiology, multi-modal information, importance context, exist, including identification ground truth. Finally, identify underexplored areas chronic connections treatments, describe promising opportunities continued