Detection of Acute Nociceptive Stimuli During General Anesthesia Using Heart Rhythm Analysis by Sparse Representations - ICISA 2017 – The 24th International Conference of the Israeli Society of Anesthesiologists held jointly with the Israeli Society of Critical Care Medicine

Maria Fridman ¹ Gal Mazor ² Tomer Golani ² Yisrael Parmet ³ Haim Berkenstadt ¹ Yonina C. Eldar ² Shai Tejman-Yarden ⁴

¹Department of Anesthesiology, Sheba Medical Center, Ramat Gan
²Department of Electrical Engineering, Technion, Israel Institute of Technology, Haifa
³Department of Management Science and Engineering, Ben Gurion University, Beer Sheva
⁴The Engineering Medical Research Lab, The Edmond J. Safra International Congenital Heart Center, Sheba Medical Center. Ramat Gan

Introduction: Recent studies demonstrated the ability of sparse recovery methodologies, to aid in heart rhythm analysis and objectively detect acute pain in conscious subjects.¹The aim of this study was to evaluate the ability of a representative sparse recovery technique, the Orthogonal Matching Pursuit algorithm (OMP) with an over-complete Fourier / Wavelet dictionary, to detect an acute nociceptive stimulus during anesthesia.

Methods: 18 adult subjects, scheduled for elective laparoscopic surgery were enrolled. Mean arterial pressure (MAP) and entropy were measured 1 minute before and 1 minute after skin incision and pneumoperitoneum. The ECG sampling started 3 minutes before till 3 minutes after the stimuli.

Results: Data from 15 patients were analyzed. Stimuli induced an increase in MAP from 67 to 76 mmHg (p<0.005), heart rate and entropy did not change significantly. ECG analysis by the OMP algorithm showed that the wavelet coefficients’ occurrence density (OD) had a significant increase during the first 60 seconds rising from 0.031-0.045 Hz to 0.122 Hz. During the second minute, the OD returned to 0.051 Hz near baseline, and later during the third minute, it rose up to 0.083 during the trocar insertion. Using logistic regression the overall OD time effect was statistically significant, f(5,82)=20 (p<0.0001).

Discussion: This study demonstrated that heart rhythm analysis using the OMP algorithm may assist in objective pain analysis during anesthesia. In particular, when using both Fourier and wavelet bases for the decomposition, the wavelet coefficients’ density is indicative of pain.

Tejman-Yarden S, et al. Med Biol Eng Comput. 2016.