Overview

 

Overview

Wearable Sensing’s dry sensors are based on ‘s leading Dry Sensor Interface (DSI) technology.
The features of the DSI technology include:

 

Dry Sensors

No Skin Preparation:

Patented electronics allow high fidelity EEG signal acquisition without the need for skin abrasion or preparation

No Gels:

Ultra-high impedance amplifiers mean that DSI sensors require no gels or fluids, and consequently do not leave messy residue. All contacts are completely dry, no sticky or wet pads anywhere! This allows for irritation-free long-term recordings, without risk of drying out!

Through-Hair Operation and Comfort:

Round-ended pinned active electrodes contact scalp through-hair and provide comfortable contact.

Uncompromising Signal Quality:

Signal quality is comparable to that obtained from conventional wet electrodes, with >90% correlation.

Artifact Resistant:

Mechanical and Engineering designs eliminate motion and electrical artifacts WITHOUT the need for advanced filtering!

(read more about DSI signal quality)

Seemlessly Integrated into

 

Rapid Set-up:

Easy-to-use headset is typically self-donned in less than 5 minutes with minimal or no adjustments needed

Even More Rapid Removal:

Easy-to-use headset is instantly taken off, and does not necessitate washing head afterwards

Ambulatory and Wireless:

Fully integrated batteries, amplifiers, digitizers, and BlueTooth transmitter allow ambulatory use

10/20 Locations:

Headsets are designed to automatically and reproducibly position sensors according to the 10/20 International System on a wide range of head-sizes.

Comfortable:

Padding and spring loaded designs provide for long-term wearable comfort unrivaled in the field of EEG. Headsets do not require chin-straps!

Analytical Algorithms

  • Quantitative Cognitive State Assessment
  • Adaptable to Various Cognitive States
  • Validated for Mental Workload, Engagement, Fatigue
  • Uses EEG, EMG, ECG, and EOG data
  • Easy to Interpret Output
  • Fast Learning Algorithm
  • Real-Time and Off-Line Analysis

Cognitive Monitoring

  • Brain-Computer Interfaces (BCI)
  • Psychological Research
  • Cognitive Workload Monitoring
  • Augmented Cognition
  • NeuroErgonomics
  • NeuroFeedback
  • NeuroMarketing
  • Biometric Analysis
  • Stress Assessment
  • Lie Detection
  • Peak Performance Training
  • Guided Meditation

 

Electroencephalography (EEG) 

Introduction

Electroencephalography (EEG) is the measurement of electric potentials at the scalp due to currents flowing through scalp tissue. The strength and distribution of currents (and therefore potentials) reflects the intensity and position of activity in the underlying neural tissue. EEG signal is measured between two electrodes, the position of which determines the recorded brain area. Multiple electrodes are typically placed in standard arrangements that cover the entire scalp and allow investigators to observe the activity of the entire brain simultaneously.
EEG is typically recorded as a time-series of potential differences, which can be evaluated visually, or analyzed spectrally, or through the use of source localization methods. The principal spectral components of EEG are divided into the following signal bands: delta (0-4 Hz), theta (4-8 Hz), alpha (8-12 Hz), beta (above 12 Hz) and gamma (above 40 Hz). Many studies have related changes in various spectral components of EEG to specific cognitive functions and clinical conditions.

Applications:

In clinical settings, continuous EEG recordings are used for monitoring sleep and anesthesia, and diagnosis of epilepsy, coma, brain death, and more recently ADHD. In research environments, EEG recordings are used in wide range of applications, including: neuroscience and cognitive psychology research into understanding brain function; Brain-Computer Interfaces to allow control of computers or machines from neural signals; neurofeedback where users are able to train their brains to generate specific activity patterns; neuromarketing where marketing companies seek to tap directly into brain signals to assess subjects’ engagement levels; neuroergonomics where researchers quantify mental workload under various strains; gaming where EEG signals are used to control computer games and toys; etc.

Check out Wikipedia’s entry on EEG for a more detailed introduction to EEG.

Data Quality

Dry Sensor Data Quality

Comparison of Dry Sensor and Wet Electrode Signal Quality

Practical sensing of biopotentials such as the electroencephalogram (EEG) in operational settings has been severely limited by the need for skin preparation and conductive electrolytes at the skin-sensor interface. Another seldom-noted problem has been the need for a low impedance connection from the body to ground for cancellation of common-mode noise voltages. In this report we describe EEG results acquired using EEG hardware based upon dry contact electrode technology, and which uses a proprietary common-mode follower (CMF) which allows a dry electrode to be used for the ground. This article presents results auditory evoked potential measurements using Wearable Sensing’s DSI-24 system simultaneously with conventional (wet) EEG electrodes. The correlations between wet and dry electrodes (averaged over 3 subjects) were 93.6% and 95.7% for F3-P3 and F4-P4, respectively.

Methods

Experimental Design:

A total of 3 subjects were selected for testing of QUASAR’s EEG hardware, according to an IRB-approved protocol. The auditory ERP task used a tone generation routine (200 tones on PC speakers, average interval 2 seconds) to stimulate ERP signals. A trigger signal was output for each tone on a single line on the parallel port of the PC to the trigger inputs of EEG hardware.

EEG Hardware:

Dry) Subjects wore Wearable Sensing’s DSI-24 EEG headset, which includes integrated dry electrode biosensors positioned at approximate standard International 10/20 electrode locations.
Wet) Wet electrode measurements were acquired using Ag/AgCl EEG electrode cups filled with Grass EC2 conductive EEG paste (Astro-Med, West Warwick, RI) and attached to sites on the subject’s scalp. The electrode sites were cleaned with alcohol to remove fats and then abraded with NuPrep (Weaver & Co., Aurora, CO). Wet electrode signals were acquired using a commercial passive wet electrode EEG amplifier that has 24-bit resolution on 16 channels of EEG and a single trigger input.The wet electrodes were positioned at F1, F5, F2, F6, P1, P2, P5, P6 electrode sites and the ground and reference electrodes were placed at the right earlobe and pinna, respectively.

Data Analysis:

The F3-P3 and F4-P4 vectors were digitally calculated from the the DSI-24 sensors. The equivalent signals for the wet electrodes were approximated by combining the wet electrode signals thus:

F3-P3 = (F1+F5)/2 – (P1+P5)/2 and F4-P4 = (F2+F6)/2 – (P2+P6)/2

Wet and Dry F3-P3 & F4-P4 signals were digitally filtered using Infinite Impulse Response (IIR) notch filters, and then bandpass filtered in a 1-40Hz bandwidth (-3dB). ERP epochs were obtained by taking an interval [-0.5s, +0.5s] around each trigger. Epochs in which the filtered signal magnitude exceeded 50 microV were rejected. The sample correlation coefficient was then calculated between the average dry electrode ERP and average wet electrode ERP signals.

Results

The results for all three subjects are presented in the Figures to the right, which plot the average ERP signals in the interval from 500ms preceding the trigger to 500ms following a trigger. Correlations between wet and dry electrodes (averaged across 3 subjects) for the intervals shown are 93.6% and 95.7% for F3-P3 and F4-P4, respectively.

In addition, average signal to noise ratios (SNRs) for ERP amplitude over pre-trigger noise RMS voltage across 3 subjects and vectors were 11.8 +/- 5.5 and 12.6 +/-2.2 for dry and wet recordings respectively, indicating equivalent SNR.

Discussion

Simultaneous measurements of ERP signals using dry electrode and wet electrodes excellent conservation of signal morphology between signals obtained from wet and dry electrodes; both in the pre-trigger “noise” segment, and in the N100-P200 ERP component. This is evident both in a visual inspection of the traces presented in the illustrative Figures, and also by the fact that the correlation values exceed 90% for both anterior-posterior ERP signals and that the SNRs for both electrode technologies are equivalent.

For additional Information, please see our full report (322KB pdf) and the associated data files (15MB zip).

Auditory Evoked Potential Responses


Average auditory ERP signals of 200 trials for dry electrodes (red) and wet electrodes (blue), for 1 subject at 2 different vectors (F3-P3, Top, and F4-P4, Bottom).

Cognitive States

Cognitive States

EEG based Cognitive Assessment

For a wide range of applications, it is cognitive information that is actually sought in a non-invasive and real-time fashion. Being a measure of brain activity, EEG spectral changes have been used for accurate estimation of alertness and cognitive workload, (Makeig and Jung, 1995; Pope et al., 1995) and cognitive fatigue. (Trejo, 2004) Furthermore, a number of studies have reported that theta is related to increases in attention, workload, memory load, and working memory performance, and that a large increase in alpha EEG precedes dozing off during a simple visual task. (Torsvall and Akerstedt, 1988) EEG has been used to monitor the progress of trainees through skill levels or identify indices of skill acquisition. One group reported an increase in event-related alpha power that correlated with amount of practice at a shooting task and suggested that it reflected a decrease in cortical activity associated with reduced effort required with expertise. (Kerick et al., 2004) Another group observed lower coherence associated with less cortico-cortical communication in expert marksmen compared to skilled shooters, and attributed this difference to decreased involvement of cognition with expertise. (Deeny et al., 2003)

QStates:

Over the past couple of years, QUASAR has developed QStates, a software package that uses quantitative EEG and heart rate variability data for assessment of cognitive and physiological state. This learning algorithm first calculates several thousand spectral EEG features, then a Partial Least Squares algorithm uses the most salient of these features as inputs for setting weights of cognitive models based on the data collected during calibration runs of defined tasks (e.g. Easy vs. Hard Tasks or Happy vs Sad Tasks). Training models requires as little as one minute of EEG data for each state (easy vs. hard), and is computationally expedient. EEG data collected during experiments are then classified with these trained models to produce real-time cognitive state measures whose output ranges from 1 to 100 representing the probability of being in one state or the other. Many efforts at developing cognitive gauges have attempted to produce universal gauges that work for all individuals in order to produce “ready to go” systems. QUASAR’s rapidly training algorithms allow for expedient calibration within minutes. These models typically produce average classifications accuracies >90%. Furthermore, the models’ outputs track task difficulty reliably, correctly interpolating cognitive workload for tasks of intermediate difficulty compared to those used for training.

Wearable Sensing is proud to offer QUASAR’s QStates among its product line. QStates interfaces seamlessly with DSI-Streamer and the various DSI systems.

QStates’ Cognitive Workload Performance.


Average classification accuracy of models for mental workload, engagement and fatigue across 18 subjects


Average mental workload model output on 18 subjects across varying task difficulty (avg ± std).

 

Video


Download

Download

Contact us for access to download files

User Manuals

Current Software

  • DSI API v.1.13.1 (869 Kb, released 08/02/2016)
    • Application Programming Interface (API) for interfacing DSI systems with custom or 3rd party software
  • DSI-2000 v.1.1.0 (36.9MB, beta released 07/20/2015)
    • Demonstration release is intented as an introduction to the BCI2000 interface module also contributed here to the BCI2000 community by Dr. Jeremy Hill
  • MATLAB DSI-Streamer Import v.1.0 (3 KB, beta released 03/19/2015)
    • 1.MATLAB scripts allows both DSI-Streamer.csv data files, and reading DSI-Streamer data in real time