A consensus statement on detection of hippocampal sharp wave ripples and differentiation from other fast oscillations

Abstract

Decades of rodent research have established the role of hippocampal sharp wave ripples (SPW-Rs) in consolidating and guiding experience. More recently, intracranial recordings in humans have suggested their role in episodic and semantic memory. Yet, common standards for recording, detection, and reporting do not exist. Here, we outline the methodological challenges involved in detecting ripple events and offer practical recommendations to improve separation from other high-frequency oscillations. We argue that shared experimental, detection, and reporting standards will provide a solid foundation for future translational discovery.

Fig. 1. Depth profile of SPW-Rs in the hippocampal CA1-dentate axis.

a Recording with a 6-shank, 96-site linear silicon probe spanning hippocampal regions and layers in a rat (5-shanks are shown, each with 16 sites with 100 µm vertical separation). Average current source density (CSD, color) maps and superimposed LFP traces of SPW-R events (100 ms, gray) from all recording sites. Asterisk indicates reference site. Note negative sharp waves and sinks (blue) in the stratum radiatum of CA1 and CA3 and the inner molecular layer of the dentate gyrus. b Same as in (a) but the maps were constructed from the filtered signal (50–250 Hz; 30 ms long traces). Red, source; blue, sink. Reproduced from ref. , CC BY-NC-SA 3.0 (https://creativecommons.org/licenses/by-nc-sa/3.0/).

Fig. 2. Spikes from groups of distant neurons contribute substantially to fast LFP oscillations.

A Histograms of extracellular spikes (top right) extracellular voltages along the CA1 stratum oriens–stratum radiatum axis in a rhythmically bursting population with ~6% of the population firing in each 10 ms interval. Spike bursts recur periodically at 150 Hz and have a Gaussian shape. The locations of neurons that spike during one 6.7 ms ripple period are indicated by triangles in a top-down view of the pyramidal layer (left), with colors indicating the 50µm-wide ring from which the spikes originate. Voltage traces are colored correspondingly, with contributions from each ring of cells adding cumulatively from the outside in. The colors in the histograms and current traces correspond to the cumulative contribution of the neurons in the ring. B Averaged power spectra of the CA1 stratum pyramidale traces from each individual ring. The insets indicate the proportions of the total voltage power at 150 Hz generated by each ring- or disk-shaped subpopulation (i.e., the peak values of the power spectra, normalized by the power at 150 Hz in the full population). Reproduced from ref. , CC BY-NC-SA 3.0 (https://creativecommons.org/licenses/by-nc-sa/3.0/).

Fig. 3. Trial-to-trial SPW-R rate correlations within and across the hemispheres.

A SPW-R rates computed throughout the entire task (20 patients). Inset: distribution of SPW-R rates across electrodes. B Coincident activation of SPW-Rs as a function of electrode (0.86 mm in diameter) distance (intercontact distance = 4–5 mm). C Trial-to-trail correlation between SPW-R rates during memory retrieval trials. D Correlation coefficients between the contralateral sites were much weaker than across electrode pairs in the same hippocampus/subiculum. Reproduced from ref. with permission, Elsevier.

Fig. 4. Relationship between cortical ripple amplitude and local spiking.

A Locations of the microelectrode arrays with respect to four nearby iEEG channels in one participant (bottom left). Right, Intraoperative photo of implanted array in the anterior temporal lobe before and after placement of an iEEG grid over the it. Bottom, Schematic of scalp, skull and cortex with respect to one iEEG channel on the cortical surface and one array in cortex. B 1500 ms window of 1–200 Hz iEEG signal (black), 80–120 Hz band iEEG signal (blue), 80–120 Hz band LFP signals across all MEA electrodes (purple), and raster plot for sorted units (red). Reprinted from ref. .

Fig. 5. SPW-Rs aligned to verbal recall for three different detection methods.

Human intracranial hippocampal CA1 recordings were taken while patients (n = 96) performed a free recall task from a 12-word list (from ref. 109). Recalls were split into the first recall and the remaining (≥2nd) recalls from each list. Ripples were detected using three different published methods (refs. 34, 36, 111) and peri-vocalization time histograms were averaged across trials pooled for all patients using 100 ms bins and a 5-point triangle smooth. While the rise in ripples before recall vocalization for ≥2nd recalls compared to 1st recalls is statistically different for all three detectors, the detected ripple rates vary several-fold depending on the detection method. Figure courtesy of John Sakon.

Fig. 6. Relationship between Ach levels in the hippocampus and SPW-R/gamma power.

a Power spectrum (0–400 Hz) and time-resolved power spectrum (40–400 Hz) of the LFP recorded from the CA1 pyramidal layer of a mouse, centered at the natural fluctuations of Ach levels (troughs and peaks, respectively). b Same as in b but during waking. Note the absence of ripples (>100 Hz) during Ach peaks, whereas highest gamma power (40–120 Hz) is present at the highest levels of Ach release. Note also the different calibration of the power panels and averaged Ach signal. Based on >50 average epochs. c Average cross-frequency power comodulogram of the LFP from the CA1 pyramidal layer in a macaque. Note the inverse correlation between ripple and beta/gamma (20–80 Hz) frequency bands. a, b Reproduced from ref. , and c reproduced from ref. , CC BY-ND 4.0 (https://creativecommons.org/licenses/by-nd/4.0/).

Fig. 7. A machine learning approach to pattern classification.

a The recording is segmented into (possibly overlapping) snippets short enough to contain at most one event. b Fourier transform of the event. c Spectral features extracted from Fourier/wavelet transform, followed by postprocessing steps. d Clustering is performed on the resulting features. Figure courtesy of Zhenrui Liao.