HIAMP

RECORD HIGH-FREQUENCY OSCILLATIONS

High-frequency oscillations (HFOs) can be recorded with the Electrocorticogram (ECoG) with implanted electrode grids in epilepsy patients. The HFOs occur in frequencies between 80-500 Hz. Hence, recording HFOs requires a high-performance biosignal amplifier such as g.HIamp with a high sampling frequency and a very good signal-to-noise ratio and resolution.

HFOs can be found in normal brain structures, but also in pathological networks, and mapping these neural networks is important for pre-surgical planning in epilepsy patients. Therefore, brain surgeons need to distinguish physiological from pathological HFOs. Recently, this was realized by the University of Houston with unsupervised machine learning techniques.

This technology can detect the repetitive waveforms of pathological HFOs within a few minutes, what is an important step for a valid clinical biomarker.

ACCURACY & DATA QUALITY

The amplifier drives each ADC at 614,4 kHz, which is much higher than the required sampling frequency. Then, the floating-point DSP internally performs the oversampling and averages samples to increase the signal-to-noise ratio. The floating-point DSP also performs real-time bandpass filtering and notch filtering of the data. Several hundred different bandpass filters are predefined.

Bipolar derivations can be calculated by the DSP to work with a very high CMRR. g.HIamp uses two additional co-processors for a steep anti-aliasing filter that guarantees a superior signal-to-noise ratio.

All these features together make it possible to record even high-gamma activity in the kHz range! The amplifier uses 256 ADC for the 256 channels, and hence all signals are sampled exactly at the same time point to avoid any time delay between channels. This is especially important for brain mapping procedures.

INPUT CHANNEL PROPERTIES

g.HIamp uses wide-range DC-coupled amplifier technology in combination with 24-bit sampling. The result is an input voltage of +/- 340 mV with a resolution of 85,7 nV! This means that every electrophysiological signal can be recorded directly, without additional hardware.

Neither high electrode offset voltage nor large artifacts resulting from electrical or magnetic stimulation will saturate the amplifier inputs. This feature is important for various artifact treatment and correction algorithms.

The use of digital filters avoids hardware related variations between channels.

CHANNEL COUNT

g.HIamp can be ordered with 16, 32, 80, 144 or 256 channels. The 80 channel version comes with one 64 and one 16 channel connector box. The 144 version comes with two 64 channels and one 16 channel connector box, and the 256 channel version comes with four 64 channel boxes.

If the multi-device toolbox from g.HIsys is used up to 1024 channels can be recorded.

SKIN-ELECTRODE IMPEDANCE

g.HIamp uses a new principle for impedance measurement that can determine the skin-electrode impedance for both passive and active electrodes. It works for both gel or dry electrodes provided by g.tec, and even for ECoG grids.

The impedance values are color coded, and all 256 channels are shown in one window, which is updated every few seconds! The impedance check can also beep if an electrode was successfully mounted. This allows very fast assembly of electrodes during real-time impedance control.

SYNCHRONIZED RECORDINGS

Combined with fNIRS, g.HIamp enables synchronized acquisition of EEG and fNIRS signals within a single research platform. Researchers can integrate active  g.LADYbird EEG electrodes, active g.SCARABEO EEG electrodes, or hybrid g.SAHARA EEG electrodes with up to 8 fNIRS sensors while maintaining precise timing and real-time data access.

The g.HIamp ecosystem supports multimodal neuroscience, machine learning workflows, Brain-Computer Interfaces, and advanced research applications where combining electrophysiological and hemodynamic measurements provides deeper insight into brain function.

TECHNOLOGY FOR REAL-WORLD NEUROSCIENCE

Studying human interaction in natural environments requires more than high-quality signal acquisition. The g.HIamp ecosystem combines high-density recordings, real-time processing with g.HIsys, multimodal synchronization, and scalable research infrastructure to support complex experiments involving multiple participants and data streams.

Researchers can synchronize EEG, EMG, ECG, eye tracking, motion capture, physiological sensors, and external devices while analyzing neural activity in real time. From hyperscanning and social neuroscience to Brain-Computer Interfaces and multimodal research, the g.HIamp ecosystem provides the flexibility, interoperability, and real-time capabilities required for next-generation neuroscience research.

SCIENTIFIC IMPACT & VALIDATION

g.HIamp is trusted by researchers worldwide for studies that demand exceptional signal quality, precise timing, real-time processing, and seamless integration with advanced neuroscience technologies. The platform has been used in peer-reviewed research spanning epilepsy surgery, functional brain mapping, Brain-Computer Interfaces (BCIs), TMS-EEG, EEG-fNIRS, high-density ECoG, multimodal neuroimaging, machine learning, and closed-loop neuroscience applications.

Published in leading journals including Nature Communications, Scientific Reports, Brain, Frontiers, Clinical Neurophysiology, Journal of Neuroscience Methods, and World Neurosurgery, these studies demonstrate how g.HIamp enables researchers to measure cortical excitability, map eloquent brain regions, decode neural activity, investigate brain connectivity, and develop next-generation neurotechnology solutions. More than a biosignal amplifier, g.HIamp serves as a complete neuroscience platform that transforms high-quality neural recordings into meaningful scientific discoveries.

LIST OF PUBLICATIONS

KINETIC DECODING FOR NEUROPROSTHETICS

Also from Nuri Firat Ince’s group (University of Houston) with MD Anderson Cancer Center, this study tackled a gap in BCI research: decoding grasp force, not just movement kinematics. Four patients undergoing awake craniotomy performed sustained isometric hand grasps while high-density ECoG (128–192 channels) was streamed through g.HIamp at 2.4 kHz, synchronized via g.HIsys/Simulink with forearm EMG, grip-force, and video on a shared timebase — a real-time, multimodal-synchronization workflow essential for intraoperative recording. Low-frequency desynchronization and high-gamma synchronization appeared at grasp onset and offset but nearly vanished during steady, sustained grip, and the high-gamma response tracked the rate of change of force (“yank”: r ≈ 0.79) far more closely than raw force (r ≈ 0.31). Published in Frontiers in Neuroscience.

READ THE STUDY

INTEROCEPTIVE-EXTEROCEPTIVE BRAIN RESEARCH

Researchers at the University of Lausanne (CHUV/UNIL) used g.HIamp to test whether the brain predicts external sounds based on cardiac timing. Tone sequences were played either locked to participants’ heartbeats or at random intervals, with 63-channel EEG, ECG, and EOG recorded simultaneously at 1200 Hz. A custom MATLAB Simulink script triggered each sound online within milliseconds of the detected heartbeat — showcasing g.HIamp’s real-time DSP and g.HIsys Simulink integration for closed-loop paradigms. Omitted sounds evoked a distinct surprise response 158–270 ms post-heartbeat, but only in the synchronized condition — the first evidence that the brain uses cardiac timing to predict external events.

READ THE STUDY

DISTINGUISHING PATHOLOGICAL FROM PHYSIOLOGICAL HFOs

A multi-center team led by Nuri Firat Ince (University of Houston), with Istanbul University, Texas Children’s Hospital, and MD Anderson Cancer Center, recorded intracranial EEG from 13 epilepsy patients and 5 controls using g.HIamp, averaging 73 electrodes per patient. Sampling at 2-2.4 kHz with 24-bit resolution preserved the subtle waveform shapes needed to feed an unsupervised clustering pipeline (GMM + DBSCAN) across 16+ hours of recordings. The team found that HFOs inside the seizure onset zone repeat with highly stereotyped waveforms, unlike the irregular HFOs from healthy functional cortex – the most tightly clustered HFOs localized the seizure onset zone with 100% specificity, entirely within the surgical resection volume in patients with favorable outcomes.

READ THE STUDY