GSR-Based Auto-Monitoring of Pain Initiation/Elimination Time Using Nonlinear Dynamic Model

The administration of pain-relieving drugs requires information on the actual existence and severity of pain. This work focuses on the patient's verbal communication independent identification of the instant of generation and extinction of pain sensation using galvanic skin response (GSR). Alteration of GSR during different stages of pain was recorded along with no pain (NP) condition. With the application of pain stimuli, the nature of GSR becomes chaotic which is analyzed in the phase space reconstruction (PSR) domain. Further, time absolute integral (TAI) and Z -score-based analysis clearly provide the pain onset and offset time (POOT). Probably for the first time, POOT detection was studied using the GSR signal and PSR technique. Four pain states were considered for the study. The absolute average error of the proposed method is calculated as 1.7 and 1.1 s for the transition instant of NP to any of the pain sensations and any pain to NP sensation, respectively, considering 85 subjects. Further analysis is presented in the result section. The present method explains a model to detect POOT using GSR. The zero-crossing analysis makes the measurement robust and easy to detect POOT. Actual estimation of POOT may significantly enhance automatic drug administration management of patients undergoing exposure to pain. This can reduce the overdosing risk of opioids to reduce pain.