Real-time Threat Detection
What do you get when you mix equal parts neuroscience, computer science, and ingenuity? You get an innovative line of research and technology from a team of scientists at Science Applications International Corporation (SAIC). The team, led by SAIC Chief Scientist Dr. Laurie Gibson, has developed a sophisticated real-time monitoring system to detect threats in the environment using neural signals.
Funded under DARPA’s Cognitive Technology Threat Warning System, Gibson and her colleagues have developed a system that allows intelligence analysts and soldiers to rapidly prioritize large volumes of imagery based on models of their own brain signals. The “neural threat response model” is built from EEG recorded with a 128-channel Geodesic EEG System 300 as the analysts view images containing both stationary and moving targets.
Because the brain's visual and threat recognition systems can detect relevant objects more quickly than the mind registers them cognitively, the SAIC system is capable of presenting thousands of images that cover hundreds of miles of terrain in a matter of minutes. The neural threat response model, based on the analysts’ subconscious recognition of relevant objects, allows the system to efficiently prioritize images in real-time for subsequent review.
According to Gibson, the individual neural threat response models are so robust there is no need to retrain or update them. “The models we build are good for months and years,” she said. “For example, we still use models that were built 18 months ago with no loss in threat detection.”
Read Full Article
In this Issue:
Customer Feature Article
EGI customer builds real-time threat detection model
EGI Distributes PRANA
New biosignal analysis options
New GeoSource source estimation software features
EGI Nets in Museums
EGI donates HCGSNs to science museums in UK and Minnesota
EGI offers training courses around the world
EGI Launches New Program
Program aims to improve user experience with third party products
AdvaMed Code of Ethics
EGI meets AdvaMed's code of ethics
Where to See EGI
Conferences and events around the world
EGI offers several career opportunities
EGI Offers PRANA for Biosignal Processing
PhiTools has joined with EGI to bring PRANA’s flexible suite of tools to the dEEG research community. Import data from EGI’s Geodesic EEG System 300 and Geodesic EEG Mobile 100 EEG systems in the EDF+ format to examine sleep, alertness, biological rhythms, and stress.
PhiTools is a developer of advanced software solutions and services for biomedical engineering, physiological signal processing, data visualization, data mining, and biostatistical analysis for such fields of research as neurophysiology, psychophysiology, chronobiology, and clinical trials.
Formed by a group of PhD students from the neuroscience and computer science departments at the University of Louis Pasteur in Strasbourg, France, PhiTools was formally established in 2001. The widely used PRANA software package appears in key journals and publications. See the PhiTools References site for a list of publications featuring PRANA.
PRANA (Polygraphic Recording Analyzer) is PhiTools’ powerful software suite for human and animal psychophysiological and neurobehavioral research. PRANA offers biosignal processing and time series analysis that is customizable for specific research applications with over a dozen specialized plug-ins. PRANA tools work on a variety of digital data formats and allow you to review, interpret, annotate, and analyze a wide range of recordings, including EEG, EOG, EMG, EKG, EDA, respiration, pulse oximetry, blood pressure, temperature, activity, sound, and light intensity.
Read More About PRANA | Request Quotation
GeoSource 2.0 represents an integration of several years of development and testing by EGI’s scientists and engineers. In an effort to make dense array EEG source analysis both robust and as simple as possible, we relied on customer feedback on individual issues.
That feedback helped us improve GeoSource in a variety of ways, including optimizing it for use with dense array data (128 or 256 channels). One reason for this optimization was to address certain problems that are inherent in source analysis that can be addressed only with dense array coverage of the whole head surface.
Results and conclusions of our development and testing will be presented soon. (Contact
for a list of upcoming white papers and publications.) One important conclusion is the possible interaction between the amplitude normalization typically used with the LORETA and LAURA (often described as depth weighting) and the regularization parameter used to obtain stable solutions in the presence of noise. These conventional weighting techniques are still available in GeoSource, although their sensitivity to (and interaction with) regularization is problematic.
A significant improvement in GeoSource 2.0 is the inclusion of the sLORETA standardization method developed by Roberto Pascual-Marqui of the Key Institute. This method standardizes the variance of each dipolar source across the array of sensor positions. With whole head coverage, including the inferior regions, the 256-channel EGI recordings ensure that variance standardization is appropriate. When electrodes are located only on the top of the braincase, the variance of the dipolar fields (and thus the head surface) will not be sampled adequately to meet the sLORETA assumptions.
Another important advantage of sLORETA is its robustness to both noise and regularization. By optimizing GeoSource for the sLORETA solution method, we have been able to simplify the source model controls (by eliminating options that are no longer necessary) and to provide simpler guidelines on how to set regularization levels for a give data set to achieve stable results quickly.
For more information about the new robust and simplified features in GeoSource 2.0, contact
EGI Nets Exhibited in Science Museums
EGI was pleased to donate a 256-channel HydroCel Geodesic Sensor Net to The Science Museum in South Kensington, UK and a 128-channel newborn Net to the Science Museum in Minnesota in St. Paul.
The Minnesota Net will become a part of an exhibit on the science of early childhood development to be installed as a permanent part of the museum's Human Body Gallery in February 2011. To learn more visit the Science Museum of Minnesota.
The UK Net is part of the Who Am I? exhibit exploring what makes you, you through a fun, interactive gallery of brain science and genetics. Interested in seeing the exhibit in person or online? Go to The Science Museum.
EGI offers a number of educational courses each year.
This year's redesigned Summer School (at our world headquarters in Eugene, Oregon, USA) was a success. We offered three courses levels for basic, intermediate, and advanced students.
EGI has upcoming research training courses in China (September 12-18) and in the UK (September 27 - October 1).
For more information and to register, see our Education page.
EGI Interoperability Standards Program (ISP)
The analytic needs of neuroscience researchers cover a wide range—with the choice of the best tool often dictated by research agenda, collaboration plans, and personal preference. To accomplish their work, EGI customers count on being able to supplement Net Station’s functionality with third party or open source hardware and software. Historically, EGI maintained the interfaces with the products it directly distributed, such as E-Prime, BESA, and Smart Eye, but did not formally track the interfaces with other software commonly used by its customers. This situation has sometimes resulted in the unintentional breaking of input/output (I/O) functionality with new Net Station releases. We at EGI know this can be frustrating for the labs that are adversely affected by these transitions.
To make using GES 300 data with other software packages easier and more predictable, EGI launched the “Interoperability Standards Program” (ISP) in August 2010. Participation in the EGI ISP is open to other research software vendors and open source/freeware authors/distributors, whether or not EGI distributes the software product. The program is also open to clinical products, but with additional restrictions to meet regulatory requirements.
The ISP program includes three tiers of support:
- Tier 1 - Products EGI sells and supports
- Tier 2 - Products that adopt EGI's file format and connectivity standards, but are neither sold by nor supported by EGI
- Tier 3 - Products that customers are using with Net Station that do not adopt EGI's file format and connectivity standards
Read the Full Article
EGI's Ethics Meet High Standards
EGI is proud to announce that it has certified compliance with the AdvaMed Code of Ethics. The AdvaMed Code recognizes the special working relationship that must exist between medical device companies, researchers, and healthcare professionals to facilitate the development of innovative medical technologies. The Code seeks to assure that these important relationships conform to widely accepted standards of ethical business practice and socially responsible corporate conduct. More information on the Code is available at AdvaMed.
Where to See EGI
- September 14-18: European Sleep Research Society (ESRS), Lisbon, Portugal
- September 16-19: Resting-State Brain Connectivity (RSBC), Milwaukee, WI, USA
- September 25-28: European Federation of Neurological Societies (EFNS), Geneva, Switzerland
- September 29 - October 3: Society for Psychological Research (SPR), Portland, OR, USA
- October 7-9: DACH-Tagung, Montreux, Switzerland
- October 13-16: Child Neurology Society (CNS), Providence, RI, USA
- October 21-23:Internationales Symposium des SFB 636 (SFB), Mannheim, Germany
- October 21-24: Asian and Oceanian Epilepsy Congress (AOEC), Melbourne, Australia
- November 13-17: Society for Neuroscience (SFN), San Diego, CA, USA
Full Conference List
EGI has several job openings in our Sales and Marketing, Science, and Software departments.
Learn more about these exciting positions with dense array EEG leader, EGI.
Interested in a career with EGI? See our Employment page.