Neural synchronization: Average strength vs. temporal patterning.

https://doi.org/10.1016/j.clinph.2018.01.063 1388-2457/ 2018 International Federation of Clinical Neurophysiology. Abbreviations: EEG, electroencephalogram; UPDRS, unified Parkinson’ rating scale Excessively strong neural synchrony may contribute to the symptoms of different neurological and neuropsychiatric disorders (Uhlhaas and Singer, 2006). Thus, hypokinetic symptoms of Parkinson’s disease are associated with elevated beta-band synchrony (Kühn et al., 2009), however this association is not very consistent (Stein and Bar-Gad, 2013). One possible explanation is that this elevated synchrony is very intermittent (Park et al., 2010). The temporal variability of synchrony provides an alternative and potentially sensitive way to characterize synchronous activity (e.g., Ahn et al., 2014; Park et al., 2010). Some results (Ahn et al., 2014) suggest that temporal patterning of synchrony may be more sensitive to the changes in the underlying neural circuits (and eventually in behavior) than average synchrony strength. Here we use parkinsonian beta-band synchronization phenomena to see how the temporal patterning of synchrony may be a more sensitive correlate of behavior than the average synchrony strength. This is not a development of a new marker of parkinsonian beta activity, but an exploration of the relationship of synchrony patterning vs. synchrony strength with behavior mediated by neural synchrony. This study includes nine patients (three female) with Parkinson’s disease, age: 64.8 ± 7.6 years, disease duration: 9.8 ± 4.4 years, UPDRS motor score: 45.1 ± 8.6 OFF medication and 20.2 ± 3.9 ON medication. It’s a small, but relatively homogeneous group; we consider all subjects’ data available to us (no special selection bias). Patients had an overall improvement of 56 9% in UPDRS motor score in ON vs. OFF. All patients exhibited hypokinetic symptoms and no or only mild rest tremor. The participants provided a written informed consent and the study was approved by Indiana University IRB. EEG recordings were performed OFF medication from C3 and C4 scalp electrodes placed according to the 10:20 international system. EEG signals were amplified x5000, digitized at 20 kHz, filtered at 0–200 Hz, and saved for off-line analysis (see Fig. 1A, B). EEG signals were visually examined before analysis to confirm proper signal collection. The average duration of the recorded episodes was 166 35 s. The data were further filtered with a digital FIR filter to the beta (10–30 Hz) band (zero-phase filtering to avoid phase distortions, see, e.g., Park et al., 2010 for the details). Synchronization strength was quantified with a phase-locking measure