Abstracts

Rationale: Sleep is essential for memory consolidation in humans- both declarative memory and procedural memory, i.e., the consolidation of motor skills. It is unclear what role sleep plays in motor recovery following stroke in humans. The aim of this project is to establish biomarkers for motor recovery in slow-wave sleep microarchitecture, e.g., sleep spindles (10-16 Hz), slow-oscillations (SOs, 0.1-1 Hz), and delta waves (d waves, 1-4 Hz). Recent rodent work has shown that SOs nested within spindles track motor recovery post-stroke, but this hasn’t been confirmed in humans. In this project, we are analyzing human, post-stroke sleep EEG for correlations to motor recovery.

Methods: Overnight EEG studies 2 and 14 days following a middle cerebral artery territory stroke were obtained retrospectively. Each 30 second epoch was individually marked for NREM sleep. Spindles, Sos, and d waves were extracted from these epochs through custom routines in MATLAB. NREM-marked EEG data was referenced with respect to the mastoid electrodes, and high amplitude artifacts were removed. For d /SOs detection, signal was passed through a 0.1 Hz high pass filter and then 4 Hz low pass Butterworth filter. All positive-to-negative zero crossings, previous peaks, following troughs, and negative-to-positive zero crossings were identified. SOs were classified as waves with troughs lower than a negative threshold, and preceding peaks higher than a positive threshold. The duration between the peak and trough was between 150 ms and 500 ms. A wave was considered a d wave if the trough was lower than the negative threshold, preceded by an amplitude lower than the positive threshold, and within 500 ms. For spindle detection, EEG data was filtered using a 10 Hz high pass Butterworth filter and 16 Hz low pass Butterworth filter. A smoothed envelope of this signal was calculated using the magnitude of the Hilbert transforms with convolving by a Gaussian window (200 ms). Epochs with signal amplitude higher than the upper threshold (mean, µ + 2.5*s.d., σ) for at least one sample and amplitude higher than the lower threshold (µ + 1.5*σ) for at least 500 ms were considered Sp’s. The lower threshold was used to define the duration of the spindle. Nested SO-spindles (similar to K-complexes) were identified as spindle peaks following SO peaks within 1.5 s duration.

Results: Our preliminary results find a lower density of spindles and SO-nested spindles in the peri-infarct cortex acutely after stroke. We are currently analyzing if the there is a pattern across electrodes within the affected hemisphere. Our forthcoming analyses will comprehensively test the frequency of these events on perilesional electrodes acutely and sub-acutely post-stroke, regressing these densities to the extent of motor impairments.

Conclusions: This work shows that there is a bias in spindle and SO-nested spindles in the acute stroke cortex. Our work can inform future work on neuromodulation of sleep by identifying specific sleep features that are most important to recovery, i.e. neuroplasticity, as well as candidate stroke populations that could benefit from such therapies.

Funding: NIH grant R00NS097620, NSF award 2048231, AHA award 847486 to T.G.