Studying the electrical activity of the brain has revealed one amazing thing: our brainwaves are individualized and unique.

Changes in brain activity under different mental states can be studied with a number of methods, and the information obtained can then be used to identify features of users. More specifically, recording of brain activity can be carried out via measurements of neuronal electrical activity, wherein a common “brainwave” or signal used is the electroencephalogram (EEG). This kind of brain signal occurs continuously and is able to give information regarding an individual’s baseline mental activity as measured over a certain length of time.

Recordings of EEGs can be obtained with relatively cheap and mobile devices in comparison with other techniques of brain imaging. Examples of these techniques include functional magnetic resonance imaging (fMRI), which measures hemoglobin concentration (oxygenated and deoxygenated) post-magnetic field exposure, positron emission tomography (PET), which involves injection of a radioactive substance to track neuron metabolism, and magneto-encephalography (MEG), which detects small magnetic fields that are induced by the brain’s innate electrical currents.

Importantly, the temporal resolution of EEG-based techniques are extremely high, reaching up to the millisecond range. Hence, computational methods can easily be utilized in dynamic studies investigating the underlying principles as well as mechanisms of specific processes. For instance, manipulation and inspection of EEG data can provide information on parameters such as neurological health, attention, memory, emotions, arousal level and psychophysiological state.

Visualizing highly individualized responses as brainwaves

Although it may not appear obvious at first, how we respond to different stimuli can provide a surprising amount of information about each of us as individuals. These stimuli can span a whole spectrum, including common stimuli from clothing, faces and colors to highly memorable ones such as specific life events. To this end, responses – in every sense of the word – have long been used as a means of identification with applications in numerous sectors, not least in the field of security validation. Indeed, brain biometrics are now able to evaluate the responses of individuals to chosen stimuli in order to accurately profile their identities. This can be done by interpreting the individually unique brainwaves linked to specific thought processes.

Using brainwaves to identify people: benefits and problems

As compared to conventional biometrics such as scans of the retina and fingerprinting, brain biometrics can potentially be better in a number of ways.

A classic example would be the duplication of fingerprints through photography at an extremely high resolution to hoodwink the security system, which can be achieved covertly with ease. On the contrary, this is clearly not possible or at least much more challenging when brain activity is used as the measurement of choice with the current level of technology, which in turn translates to enhanced security when using brain biometrics.

Furthermore, many features and parameters of brain biometrics are not directly under the control of the individual user i.e. it is not possible to readily manipulate or imitate the readout of the program. This suggests that inherently, the innate mechanisms driving brain biometrics are relatively more secure than other types of biometrics. In spite of the aforementioned benefits, brain biometrics have not yet reached widespread use largely due to the fact that an identification accuracy of 100% has not been easy to achieve.

The quest to improve reliability in brain biometrics

In a recent study on brain biometrics, scientists first selected 500 images which were specifically chosen to elicit specific responses from individuals, which included words, items, foods and even the celebrity Anne Hathaway. Through the use of custom made headsets, they then examined the brain activity of 50 people who were sequentially exposed to these images. Remarkably, the reactions of the volunteers to the images was sufficiently different to the extent that the biometric system could distinguish the “brainprint” of each participant at an accuracy of 100% through his or her brain activity.

Notably, in previous attempts prior to this work, one person out of a group of 32 could be identified via an individual-specific response at an accuracy of only 97%, which would conceivably still leave a sliver of doubt regarding its robustness. It turned out that this reduction in the level of accuracy was primarily due to the fact the focus of these techniques tended towards using EEGs. The main problem with the use of EEGs was that the reliability of the identification could be compromised due to individual variability of cognitive states understandably skewing the results.

In the current study, another type of brain signal known as the event-related potential (ERP) was used to enhance the accuracy of the identification process. ERPs are short-term brain signals which are generated upon the brain’s response to certain stimuli and/or events, making for an ideal candidate parameter for capturing snapshots of information and producing unique profiles of individuals. This implies that the use of ERP as a brain biometric would be much more robust as compared to the EEG, which was indeed corroborated by the improved identification accuracy reported.

Perhaps it is also rather apt that the method described here has been dubbed the CEREBRE (Cognitive Event RElated Biometric REcognition) protocol – a clever play on the word cerebrum, a critical brain structure involved in the integration of neural and sensory functions.

More importantly, the 100% accuracy reported here represents an important breakthrough in the field, and what is exciting is that there are many possible uses for situations where high levels of security are needed. In this regard, it is definitely possible that security systems could be using brainwaves as the gold standard method of identity verification with technological advances in the future.

References

Armstrong, B., Ruiz-Blondet, M., Khalifian, N., Kurtz, K., Jin, Z., & Laszlo, S. (2015). Brainprint: Assessing the uniqueness, collectability, and permanence of a novel method for ERP biometrics Neurocomputing, 166, 59-67 DOI: 10.1016/j.neucom.2015.04.025

Campisi, P., & La Rocca, D. (2014). Brain waves for automatic biometric-based user recognition IEEE Transactions on Information Forensics and Security, 9 (5), 782-800 DOI: 10.1109/TIFS.2014.2308640

Pevzner, A., Izadi, A., Lee, D., Shahlaie, K., & Gurkoff, G. (2016). Making Waves in the Brain: What Are Oscillations, and Why Modulating Them Makes Sense for Brain Injury Frontiers in Systems Neuroscience, 10 DOI: 10.3389/fnsys.2016.00030

Ruiz-Blondet, M., Jin, Z., & Laszlo, S. (2016). CEREBRE: A Novel Method for Very High Accuracy Event-Related Potential Biometric Identification IEEE Transactions on Information Forensics and Security, 11 (7), 1618-1629 DOI: 10.1109/TIFS.2016.2543524

Yeom, S., Suk, H., & Lee, S. (2013). Person authentication from neural activity of face-specific visual self-representation Pattern Recognition, 46 (4), 1159-1169 DOI: 10.1016/j.patcog.2012.10.023

Image via byrev / Pixabay.