Recording human cortical population spikes non-invasively--An EEG tutorial.

BACKGROUND: Non-invasively recorded somatosensory high-frequency oscillations (sHFOs) evoked by electric nerve stimulation are markers of human cortical population spikes. Previously, their analysis was based on massive averaging of EEG responses. Advanced neurotechnology and optimized off-line analysis can enhance the signal-to-noise ratio of sHFOs, eventually enabling single-trial analysis. METHODS: The rationale for developing dedicated low-noise EEG technology for sHFOs is unfolded. Detailed recording procedures and tailored analysis principles are explained step-by-step. Source codes in Matlab and Python are provided as supplementary material online. RESULTS: Combining synergistic hardware and analysis improvements, evoked sHFOs at around 600 Hz ('σ-bursts') can be studied in single-trials. Additionally, optimized spatial filters increase the signal-to-noise ratio of components at about 1 kHz ('κ-bursts') enabling their detection in non-invasive surface EEG. CONCLUSIONS: sHFOs offer a unique possibility to record evoked human cortical population spikes non-invasively. The experimental approaches and algorithms presented here enable also non-specialized EEG laboratories to combine measurements of conventional low-frequency EEG with the analysis of concomitant cortical population spike responses.

Non-invasive single-trial EEG detection of evoked human neocortical population spikes.

QUESTION: Human high-frequency (>400 Hz) components of somatosensory evoked potentials (hf-SEPs), which can be recorded non-invasively at the scalp, are generated by cortical population spikes, as inferred from microelectrode recordings in non-human primates. It is a critical limitation to broader neurophysiological study of hf-SEPs in that hundreds of responses have to be averaged to detect hf-SEPs reliably. Here, we establish a framework for detecting human hf-SEPs non-invasively in single trials. METHODS: Spatio-temporal features were extracted from band-pass filtered (400-900 Hz) hf-SEPs by bilinear Common Spatio-Temporal Patterns (bCSTP) and then classified by a weighted Extreme Learning Machine (w-ELM). The effect of varying signal-to-noise ratio (SNR), number of trials, and degree of w-ELM re-weighting was characterized using surrogate data. For practical demonstration of the algorithm, median nerve hf-SEPs were recorded inside a shielded room in four subjects, spanning the hf-SEP signal-to-noise ratio characteristic for a larger population, utilizing a custom-built 29-channel low-noise EEG amplifier. RESULTS: Using surrogate data, the SNR proved to be pivotal to detect hf-SEPs in single trials efficiently, with the trade-off between sensitivity and specificity of the algorithm being obtained by the w-ELM re-weighting parameter. In practice, human hf-SEPs were detected non-invasively in single trials with a sensitivity of up to 99% and a specificity of up to 97% in two subjects, even without any recourse to knowledge of stimulus timing. Matching with the results of the surrogate data analysis, these rates dropped to 62-79% sensitivity and 18-31% specificity in two subjects with lower SNR. CONCLUSIONS: Otherwise buried in background noise, human high-frequency EEG components can be extracted from low-noise recordings. Specifically, refined supervised filter optimization and classification enables the reliable detection of single-trial hf-SEPs, representing non-invasive correlates of cortical population spikes. SIGNIFICANCE: While low-frequency EEG reflects summed postsynaptic potentials, and thereby neuronal input, we suggest that high-frequency EEG (>400 Hz) can provide non-invasive access to the unaveraged output of neuronal computation, i.e., single-trial population spike activity evoked in the responsive neuronal ensemble.

An advanced phantom study assessing the feasibility of neuronal current imaging by ultra-low-field NMR.

In ultra-low-field (ULF) NMR/MRI, a common scheme is to magnetize the sample by a polarizing field of up to hundreds of mT, after which the NMR signal, precessing in a field on the order of several µT, is detected with superconducting quantum interference devices (SQUIDs). In our ULF-NMR system, we polarize with up to 50mT and deploy a single-stage DC-SQUID current sensor with an integrated input coil which is connected to a wire-wound Nb gradiometer. We developed this system (white noise 0.50fT/√Hz) for assessing the feasibility of imaging neuronal currents by detecting their effect on the ULF-NMR signal. Magnetoencephalography investigations of evoked brain activity showed neuronal dipole moments below 50nAm. With our instrumentation, we have studied two different approaches for neuronal current imaging. In the so-called DC effect, long-lived neuronal activity shifts the Larmor frequency of the surrounding protons. An alternative strategy is to exploit fast neuronal activity as a tipping pulse. This so-called AC effect requires the proton Larmor frequency to match the frequency of the neuronal activity, which ranges from near-DC to ∼kHz. We emulated neuronal activity by means of a single dipolar source in a physical phantom, consisting of a hollow sphere filled with an aqueous solution of CuSO4 and NaCl. In these phantom studies, with physiologically relevant dipole depths, we determined resolution limits for our set-up for the AC and the DC effect of ∼10µAm and ∼50nAm, respectively. Hence, the DC effect appears to be detectable in vivo by current ULF-NMR technology.

Magnetic resonance imaging at frequencies below 1 kHz.

Within the magnetic resonance imaging (MRI) community the trend is going to higher and higher magnetic fields, ranging from 1.5 T to 7 T, corresponding to Larmor frequencies of 63.8-298 MHz. Since for high-field MRI the magnetization increases with the applied magnetic field, the signal-to-noise-ratio increases as well, thus enabling higher image resolutions. On the other hand, MRI is possible also at ultra-low magnetic fields, as was shown by different groups. The goal of our development was to reach a Larmor frequency range of the low-field MRI system corresponding to the frequency range of human brain activities ranging from near zero-frequency (near-DC) to over 1 kHz. Here, first 2D MRI images of phantoms taken at Larmor frequencies of 100 Hz and 731 Hz will be shown and discussed. These frequencies are examples of brain activity triggered by electrostimulation of the median nerve. The method will allow the magnetic fields of the brain currents to influence the magnetic resonance image, and thus lead to a direct functional imaging modality of neuronal currents.

Are brain currents detectable by means of low-field NMR? A phantom study.

A number of different methods have been developed in order to detect the spreading of neuronal currents by means of noninvasive imaging techniques. However, all of these are subjected to limitations in the temporal or spatial resolution. A new approach of neuronal current detection is based on the use of low-field nuclear magnetic resonance (LF-NMR) that records brain activity directly. In the following, we describe a phantom study in order to assess the feasibility of neuronal current detection using LF-NMR. In addition to that, necessary preliminary subject studies examining somatosensory evoked neuronal currents are presented. During the phantom study, the influences of two different neuronal time signals on (1)H-NMR signals were observed. The measurements were carried out by using a head phantom with an integrated current dipole to simulate neuronal activity. Two LF-NMR methods based on a DC and an AC (resonant) mechanism were utilized to study the feasibility of detecting both types of magnetic brain signals. Measurements were made inside an extremely magnetically shielded room by using a superconducting quantum interference device magnetometer system. The measurement principles were validated applying currents of higher intensity than those typical of the neuronal currents. Through stepwise reduction of the amplitude of the current dipole strength, the resolution limits of the two measuring procedures were found. The results indicate that it is necessary to improve the signal-to-noise ratio of the measurement system by at least a factor of 38 in order to detect typical human neuronal activity directly by means of LF-NMR. In addition to that, ways of achieving this factor are discussed.

Nuclear magnetic relaxation in water revisited.

In this study, we revisited nuclear magnetic relaxation of (1)H in water at very low Larmor frequencies that has been studied intensively in earlier years. We make use of the recently developed superconducting quantum interference device based ultra-low field NMR technique, which enables much easier access to the longitudinal spin-lattice relaxation time T(1) and the transversal spin-spin relaxation time T(2) below several kHz than traditional field cycling methods. Our data reproduce and complement the earlier results, in that they corroborate the finding of an exchange process with a correlation time of about 0.34 ms at room temperature which can be attributed to the migration of hydronium and hydroxyl ions in neutral water via hydrogen bridges. The corresponding relaxation process is driven by the interaction of the protons with (17)O and contributes to the T(1) and the T(2) relaxation rate by about 0.12 s(-1). In addition, we found evidence of a very slow exchange process at about 100 Hz that has hitherto not been reported.

Simultaneous measurements of somatosensory evoked AC and near-DC MEG signals.

Magnetoencephalography measurements of somatosensory evoked brain activity taken inside an extremely magnetically shielded room are reported. The massive low frequency shielding in combination with a high sampling rate enabled the simultaneous observation of AC and near-DC effects. Neuronal activation was achieved by repetitive electrostimulation of the right median nerve above motor threshold using repetition rates from 3 Hz to 12 Hz. Stimulation sequences lasted for 10 s and were interspersed with periods of rest of equal length. The recorded magnetic fields, inferred for the N20m and a sustained near-DC component, revealed mainly dipolar patterns with mutually rotated orientations with angles of rotation of 30° and 75°. At the start of the stimulation we observed a fast rise within 100 ms in the evoked magnetic near-DC fields for which a maximum equivalent current dipole strength of 65 nAm was obtained. The sustained fields decayed by a factor of ∼4 to a lower DC-level B(γ) with a time constant τ of order of seconds. For 12 Hz repetition rate B(γ) was decreased. We suggest that the sustained neuronal activity evoked by repetitive electrostimulation could provide a suitable scheme to realize the direct detection of DC effects of neuronal currents via low field magnetic resonance.