Normalized compression distance to measure cortico-muscular synchronization

Abstract

The neuronal functional connectivity is a complex and non-stationary phenomenon creating dynamic networks synchronization determining the brain states and needed to produce tasks. Here, as a measure that quantifies the synchronization between the neuronal electrical activity of two brain regions, we used the normalized compression distance (NCD), which is the length of the compressed file constituted by the concatenated two signals, normalized by the length of the two compressed files including each single signal. To test the NCD sensitivity to physiological properties, we used NCD to measure the cortico-muscular synchronization, a well-known mechanism to control movements, in 15 healthy volunteers during a weak handgrip. Independently of NCD compressor (Huffman or Lempel Ziv), we found out that the resulting measure is sensitive to the dominant-non dominant asymmetry when novelty management is required (p = 0.011; p = 0.007, respectively) and depends on the level of novelty when moving the non-dominant hand (p = 0.012; p = 0.024). Showing lower synchronization levels for less dexterous networks, NCD seems to be a measure able to enrich the estimate of functional two-node connectivity within the neuronal networks that control the body.

Keywords: electrophysiology; feedback; handedness; neuronal synchronization; normalized compression distance (NCD).

FIGURE 1

Experimental setting. (A) Electro-encephalography (EEG) recordings and task. The general set-up to record the EEG during the weak handgrip executed in sequences of 20 s starting with a go signal (green rectangle) and ending with a stop signal (red rectangle) intermingled by 10 s at rest. In the visual feedback “yes” (“no”) condition, a horizontal segment indicates the level of exerted pressure on the bulb by vertical oscillations (blocked). After determining handgrip maximum voluntary contraction (MVC), a rest period of at least 2 min was provided. Then, the weak isometric handgrip and rest sequences lasted about 5 min. The target level was set to 5% MVC, to minimize weariness related to the task. (B) Example of electro-myographic (EMG) acquisition during isometric contraction execution. In gray the EMG trace of opponents pollicis (OP) muscle in one representative subject, for the whole task duration, with 20 s contraction sequences intermingled by 10 s at rest. Light blue line indicates the temporal portions selected for analysis.

FIGURE 2

Variables of interest. In the four conditions of interest, representation of the functional connectivity measure obtained by normalized compression distance (NCD) from the electro-encephalography (EEG) and electro-myographic (EMG) ongoing signals. In red (left movement) and blue (right movement) the conditions with visual feedback (Yes), and in light red and light blue the conditions without (No). In the topographical representation of the 64-EEG recording channels, we highlighted those considered to estimate the bipolar derivation displaying highest cortico-muscular coherence (CMC) during the task with the contralateral hand, used as criterion to select the EEG representative.

FIGURE 3

CMncdH and CMncdLZ dependence on behavioral condition. Boxplot of CMncdH (A) and CMncdLZ (B) reporting for each subject the value in 6 estimate intervals in the four conditions. Black lines with asterisks: conditions differing for p < 0.05 (1 asterisk) and p < 0.01 (2 asterisks). In panel (C) we show for comparison single subject data of the study L’Abbate et al. (2022), where we observed a that in absence of visual feedback cortico-muscular coherence (CMC) had higher amplitudes than in presence. Note that higher CMC corresponds to lower CMncd.