MMBT-S Trigger Interface Box
New: In-Built LED to Visualise Trigger Markers and Accessible Runtime Mode Switch!Browse
Bringing you World-Leading Solutions for Research in Neurosciences
Search for Products in one of the main categories below:
Passive vs. Active Electrode - Slew-Rate Problem New
It was brought to our attention that a rumour is buzzing around among EEG researchers that there may be an inherent noise problem with active electrodes. The source of this rumour is a paper published by Sarah Laszlo et al. in 2014 (see A direct comparison of active and passive ampliﬁcation electrodes in the same ampliﬁer system). In this paper, the authors compare the performance of one particular brand (not BioSemi) active electrodes with passive electrodes, both connected to a conventional EEG amplifier (also not BioSemi). Both active and passive electrodes setups were tested with low (2 kOhm) and high (50 kOhm) electrode-to-skin impedances.
The authors find that the total noise (250 Hz bandwidth) for passive electrodes is higher with high electrode impedance, whereas the total noise with the active electrodes is equal (and nearly as low as with passive electrodes with low impedance) for both low and high electrode impedances. However, they also determine that the number of sweeps needed to achieve a reliable ERP result is higher with high electrode impedances for both the active and passive setups. Note though, that the number of required sweeps did not show significant differences between active and passive setups.
The paper starts to derail when the authors attempt to explain why in the tests with high skin-to-electrode impedances, the number of sweeps required for reliable ERP is similar for active and passive setups, instead of a lower number of sweeps for the active case as they would expect on the basis of the better total noise results. So although the active electrodes do not perform any worse than passive electrodes, they basically wonder why the performance is not better than it is.
A logical first step in analyzing the above discrepancy would have been to analyze the noise spectra (for example a relatively small band of extra noise with frequencies around the ERP waveform would explain that the total noise number is hardly influenced whereas the number of required sweeps would certainly be affected). Instead, the authors choose to introduce a hypothesis without any support by measurements or references. The authors suggest that the amplifier in the used active electrodes has a slew rate that is limited to such an extent that it causes distortion of the ERP wave, leading to a higher required number of sweeps than would be expected on the basis of the total noise figures.
It is certainly true that slew rate limitations can lead to signal distortion. The effect usually plays a role only with large (several volts) signals at high (several MHz) frequencies. Suggesting slew rate limitations as a problem for EEG signals (amplitudes in the hundreds of uV range and frequencies below a few hundred Hz) sounds far fetched.
How extreme the hypothesis actually is, can be illustrated by some numbers. The paper shows ERP results with a bandwidth of 250 Hz and an amplitude of 6 uV. The required slew rate to acquire this waveform without distortion is approx. 0.01 V/s (min SR = 2*pi*f*Vpk-pk, see for example the Wikipedia page on Slew Rate). Now consider the standard BioSemi ActiveTwo EEG system with active electrodes. At 16 kHz sample rate, this system can acquire a 3 kHz, 0.5 Vpk-pk wave without significant distortion. The complete system (active electrodes, post-amplifier, analog-to-digital converter) therefore has a slew rate of at least 10,000 V/s (or 0.01 V/us in the more usual notation). The standard BioSemi active electrodes are also applied in a high-frequency version of the ActiveTwo system with 262 kHz sample rate. The setup accurately acquires a maximal sine wave of 80 kHz, 0.5 Vpk-pk, leading to a slew rate estimate for the standard BioSemi active electrodes of at least 0.2 V/us. This is a factor of 20 million faster than in the Laszlo hypothesis. Refer to the following result for a 20 kHz, 40 mVpk-pk input signal acquired with standard BioSemi electrodes and a 262 kHz sample rate version of the ActiveTwo system (the dots indicate sample points at 3.8 us intervals). Note the fast rise and fall times of the square wave edges indicating good slew rate performance (the plot shows rise/fall times of approx 10 us, this is a limitation for step responses of the decimation filter in the ADC).
Designers of EEG systems, with or without active electrodes, all make use of a relatively small pool of quite similar Op-amps (building blocks for amplifier circuits) and ADCs provided by a handful of semiconductor manufacturers. These components all offer slew rates far beyond what is minimally required for EEG signals. I am therefore quite sure that our competitors can present results similar as found for the BioSemi setup. This is supported by the input range and bandwidth figures published by various manufacturers. While I would not hesitate to spread doubts about performance aspects of products by BioSemi's competitors, the limited slew rate argument is so unlikely for any currently available EEG system (with or without active electrodes) that I would never consider using it.
In other words: it is doubtful that the particular (non BioSemi) active electrode evaluated in the paper has a slew rate with a factor of 20 million slower than the BioSemi active electrode. Anyway, such an extremely slow slew rate would have been immediately apparent in the waveforms generated by relatively large and fast artifacts in the EEG signal, such as eyeblinks. In addition, the phenomenon would have been straight forward to measure by connecting the active electrodes to a signal generator. The paper does not indicate that any attempt was made to verify the hypothesis with measurements. Given that the particular active electrodes as tested in the paper are quite widespread under EEG researchers, it is hard to imagine that a slew rate limitation of 0.01 V/s would have remained unnoticed for years among these researchers.
The authors make things worse by stating that slew rate limitations are more severe for amplifiers "at very low output impedance". They even refer to BioSemi to support this claim. In reality, BioSemi never has made such a statement and with good reason: it simply is not true. There are numerous amplifier designs with high slew rate and low output impedance, the BioSemi active electrode as presented above is just an example.
The last sentence of the paragraph is the most bizarre. While it is, of course, true that passive electrodes do not have the slew rate limitations associated with amplifier circuitry, the authors overlook that the input stage of an EEG amplifier for passive electrodes also has high input impedance, low output impedance and slew rate limitations. In terms of amplifier noise, input and output impedances and slew rate, the only principle difference between the active and the passive setup is the location of the first amplifier stage. The statement that "slew rate [..] is simply not an issue for passive electrodes" is therefore surprising. In reality, both active electrodes and input stages for passive electrodes have the same slew rate limitations, although as outlined above, it is unlikely that any EEG system (active or passive) using currently available electronic components will present a problem in this respect. Instead, the authors suggest that slew-rate limitation is an inherent problem encountered with all active electrodes (regardless of brand and design), whereas the effect is totally eliminated in all amplifiers for passive electrodes.
To summarize: the paper presents an unlikely hypothesis that is not verified and not supported by any valid reference. The authors fail to do the simple measurements or numerical evaluation that would have readily falsified the hypothesis. Nevertheless, they find themselves qualified to apply this questionable hypothesis not only to the particular active electrode evaluated in the paper but to the active electrode design in general. In effect, they disqualify all active electrodes, including the successful BioSemi design, and confuse the scientific discussion about the merits of the active electrode principle. It is peculiar to see that such an unjustified accusation, that may affect legitimate manufacturers who put a lot of effort in improving EEG acquisition systems, can make it through the review process of Journal of Neuroscience Methods.
This is an article published by Coen Metting van Rijn, PhD, director, BioSemi, 5-December 2019
FAQ: How shall I disinfect EEG caps and electrodes and what kind of disinfection solution shall I use after each recording sessions? New
Soap is a lipophilic substance, meaning that it has the ability to chemically bind to fatty/oily substances. The weakest link of the novel SARS-CoV-2 (a.k.a. COVID-19 virus) is the lipid (fatty) bilayer membrane. Due to this fact, soap will resolve the fatty outer layer of the virus and it becomes inactive. We advise cleaning the head-caps with warm water and soap. A shampoo optimized for delicate fabric is advised, such as Ivory Shampoo (which can be purchased from our online shop: https://shop.neurospec.com/ivory-electrode-cap-shampoo-750ml), or similar. When cleaning, use a soft brush to clean and degrease the inside of the electrode holders.
According to literature, alcohol (ethanol or isopropanol) makes the virus inactive within 30 seconds. The electrodes can be submerged in alcohol without damage. Hence, submerging the electrodes for five minutes in alcohol should be the most effective way to prevent COVID-19 infections. Furthermore, the U.S. Food and Drug Administration (FDA) recommends using alcohol-based disinfectant solutions with greater than 60% ethanol or 70% isopropanol.
BESA MRI 3.0 released New
BESA MRI 3.0 has just been released and we are happy to announce the several new features to enhance your data analysis and results review. These new features include:
- Automated Boundary Element Model (BEM) calculation for use in BESA Research source modelling
- Sufficient automated algorithm for placing electrodes recorded in standard 10-10 positions onto individual head surface reconstructed from MRI
- User-configurable multi-slice view of MRI images
- Several brain atlases available for display on structural MRI in a single view or multi-slice view
- Display BESA Research discrete source analysis solutions in individual MRI with confidence limit of source location
- Support for lossless .jpg compressed DICOM files
- Improvements in co-registration visualization and user interaction
The latest BESA MRI 3.0 runs in combination with BESA Research 7.1 and 7.0. Individual head models (BEM, FEM) generated with BESA MRI 3.0 can be imported from BESA Research 7.x for EEG or MEG source analysis. EEG / MEG data can be co-registered with individual MRI data using individually digitized electrode positions or head shape models.
Get the latest BESA MRI 3.0 license directly from the NEUROSPEC online shop. Visit our online store today to get the latest version of BESA MRI today!
Analysing Timing Accuracy of the ActiveTwo (BioSemi) New
Manufacturers of EEG systems all face similar challenges when it comes to designing hardware for high-quality and time accurate EEG measurements. High-quality components, good data recording techniques and the right selection of tools is crucial to acquiring reliable and repeatable results for a wide array of life science applications including Psychophysiology, Exercise Physiology, Sleep Studies, and more.
In this video, we will explain and discuss how to precisely measure and analyse delays of your EEG system. We will briefly demonstrate a typical EEG setup with stimulation presentation computer running a VEP (visually evoked potential), simultaneously sending triggers through the MMBT-S to the receiving EEG system (ActiveTwo, BioSemi).
You Will Learn:
- How to measure delays between incoming triggers (on the USB receiver) and trigger onset (ERGO input on the AD-Box)
- A possible EEG system (ActiveTwo, BioSemi) setup with a stimulus presentation computer running a VEP (brief overview)
- Sending triggers from the software Presentation, through the MMBT-S to the ActiveTwo EEG system (brief overview)
- How to record an EEG datafile with triggers in ActiView (brief overview)
- How to read a recorded EEG datafile with triggers in the BDF reader software from BioSemi
- How to utilize the BDF reader software from BioSemi
Warning: this setup is for demonstration purposes only and must not be replicated!
How Paraplegics Could Learn To Walk Again New
See the DSI-24 by Wearable Sensing in Action. In this interesting presentation titled "Implant in the Brain - How paraplegics could learn to walk again" by SRF Einstein.
BESA Statistics 2.1 - Released! New
BESA Statistics 2.1 now includes new dedicated workflows to perform t-test, single-factor ANOVA and correlation analysis of your data using the parameter-free cluster permutation statistics which solves the multiple-test problem. Furthermore, BESA Statistics supports additional data types to ensure enhanced time-frequency analysis and connectivity analysis supported.
Highlights of the latest BESA Statistics release include:
- Support of data type Connectivity. This enables direct import of results obtained in BESA Connectivity for group statistics on connectivity results in sensor space or source space.
- Configurable slice view for Image data to displays sequences in one of three available orthogonal orientations.
- New colour scheme: BESA White.
- Several new colour maps are available.
- Data values are now displayed on mouse-over in the detail windows.
- Time-frequency data stored by BESA Connectivity with wavelet analysis can now be read with the correct (logarithmic) frequency spacing.
- Single-trial time-frequency data can now be read in the t-test workflow (*.tfcs data format).
- There is no upper limit on the number of data files imported into the workflow.
- New image export format available: SVG.
- Screenshots and cluster summary results can now be copied to the clipboard using the right mouse popup menu.
- High-DPI displays are supported.
Please check the Update History for details.
BESA Statistics provides optimized, user-guided workflows for the cross-subject analysis of EEG / MEG data. The statistical method used is parameter-free permutation testing on the basis of Student’s t-tests (Maris, E. and Oostenveld, R., 2007), F-tests (for ANOVA/ANCOVA), and correlations. Statistical values computed in BESA Statistics 2.1 can be directly used for scientific reports. No further analysis in other programs is needed. All results are visualized and can be directly used for publications.
BESA Statistics 2.1 integrates pre-analyzed data from BESA Research and can also process data from other software packages as long as they conform to the BESA Statistics file format. The BrainVision Analyzer 2 native data format is supported for time and time-frequency data.
BESA Statistics will automatically identify clusters in time, and if applicable frequency and space where data of the input groups/conditions are not interchangeable, i.e. where the null-hypothesis must be rejected. Results are considered corrected for multiple comparisons as only those clusters will be identified that have higher cluster values than 95% of all clusters derived by random permutation of data. Thus, results obtained by BESA Statistics are objective and robust.
For ANOVA/ANCOVA analysis, an additional non-parametric post-hoc Scheffe’s test is computed to determine, which pairwise comparison(s) were responsible for the group/condition main effect. A Bonferroni-Holm correction for multiple comparisons of the different pairwise combinations is applied subsequently.
AS4SAN 2020 UPDATE New
With the unfortunate global situation concerning the Covid-19, we are sad to inform you that the AS4SAN committee has decided to postpone the conference until further notice. We are all in this situation together and hope that you stay safe. More updates will come as soon as we know more.
For more information keep an eye on the homepage of the AS4SAN conference.
New Wearable Sensing Webinar (14.04.2021) New
Join us for this new Wearable Sensing webinar. Wearable Sensing will introduce its new innovative EEG+fNIR system. The DSI-fNIR integrates Dry Electrode EEG sensors with functional Near- Infrared (fNIR) sensors into a single headset. There are two sessions so register for your session of choice now.
The non-intrusive wireless DSI-fNIR headset is designed for synchronized recording of the brain’s electrical activity and its hemodynamic response or blood-oxygen-level-dependent (BOLD) response in ambulatory real- and virtual environments. The DSI-fNIR's sensors are arranged to allow simultaneous and superimposed recordings at locations distributed across the scalp. The system produces raw data, and a machine learning algorithm fuses both signals for cognitive state classification. The webinar will describe this technology and briefly present a few applications that highlight the advantages of such bimodal EEG+fNIR sensing.
SOME OF OUR RANGE OF PRODUCTS
BIOSEMI - ActiveTwo
BIOSEMI - bringing EEG and ERP to a new level with the original Active Electrode and ActiveTwo EEG Amplifier.
BESA - Research 7.0
BESA - the leading innovators in digital EEG, MEG and MRI software for research and clinical applications.
SHIMADZU - LABNIRS
SHIMADZU - Next-Generation Optical Brain-Function Imaging with functional near-infrared spectroscopy (fNIRS)