Single-sweep spectral analysis of contact heat evoked potentials: a novel approach to identify altered cortical processing after morphine treatment

Abstract

Aims: The cortical response to nociceptive thermal stimuli recorded as contact heat evoked potentials (CHEPs) may be altered by morphine. However, previous studies have averaged CHEPs over multiple stimuli, which are confounded by jitter between sweeps. Thus, the aim was to assess single-sweep characteristics to identify alterations induced by morphine.

Methods: In a crossover study 15 single-sweep CHEPs were analyzed from 62 electroencephalography electrodes in 26 healthy volunteers before and after administration of morphine or placebo. Each sweep was decomposed by a continuous wavelet transform to obtain normalized spectral indices in the delta (0.5-4 Hz), theta (4-8 Hz), alpha (8-12 Hz), beta (12-32 Hz) and gamma (32-80 Hz) bands. The average distribution over all sweeps and channels was calculated for the four recordings for each volunteer, and the two recordings before treatments were assessed for reproducibility. Baseline corrected spectral indices after morphine and placebo treatments were compared to identify alterations induced by morphine.

Results: Reproducibility between baseline CHEPs was demonstrated. As compared with placebo, morphine decreased the spectral indices in the delta and theta bands by 13% (P = 0.04) and 9% (P = 0.007), while the beta and gamma bands were increased by 10% (P = 0.006) and 24% (P = 0.04).

Conclusion: The decreases in the delta and theta band are suggested to represent a decrease in the pain specific morphology of the CHEPs, which indicates a diminished pain response after morphine administration. Hence, assessment of spectral indices in single-sweep CHEPs can be used to study cortical mechanisms induced by morphine treatment.

Trial registration: ClinicalTrials.gov NCT01245244.

Keywords: contact heat; electroencephalography; evoked brain potentials; morphine; spectral analysis.

© 2015 The British Pharmacological Society.

Figures

Figure 1

Schematic illustration of the spectral analysis procedure of the EEG signals. For each subject, each condition (before and after morphine and placebo administration), each channel and each single-sweep, the continuous wavelet transform was preformed and the normalized spectral power distribution was calculated. The spectral power distribution was averaged over all sweeps to obtain one estimate of the spectral indices for each subject, condition and channel. An average over all channels was used to calculate the ratio after administration : before administration for both the morphine and placebo sessions. The average sweep distribution was used to calculate grand mean (GM) for all subjects for each condition to create topography maps and statistical analysis of the ratio after administration : before administration was performed for all subjects

Figure 2

CHEPs from a recording before drug including 15 stimuli from a representative subject for the Cz channel. The vertical lines show the time interval of interest for the spectral analysis. (A) The 15 single-sweeps (black) and the corresponding average evoked potential (red). The two main peaks (N2 and P2) of the average evoked potential are illustrated and due to the jitter, a large portion of the signal information is lost and the average amplitude is low. (B) Amplitude intensities for all 15 single-sweeps are represented by a colour code to visualize the variability between sweeps. Phase jitter and amplitude variation is present, but not in a dynamic way as would be the case of, for example, habituation

Figure 3

(A) A representative sweep for the Cz channel in the time domain and (B) a selected part of the time–frequency coefficients squared to obtain the power distribution. The vertical lines show the time interval of interest for the spectral analysis. The maximum power is found in the theta band (red colour) within the interval of interest. The frequency range is limited to 20 Hz in the figure as the pain specific morphology of the N2-P2 complex of the evoked potential is also located in the theta band (4.0–8.0 Hz)

Figure 4

Grand mean topography plot of spectral indices in all analyzed bands before drug administration

Figure 5

Topography plots of the altered spectral indices after treatment compared with before treatment in all analyzed bands

Figure 6

Mean ratios of spectral indices over all channels for all analyzed bands (mean ± SEM). *significant difference between morphine and placebo treatment, P < 0.05. , morphine; , placebo