Brain-Monitoring Headset for Communication | Cognixion

The Challenge of Communication with Motor Impairments

For individuals grappling with significant motor impairments, the inability to communicate effectively represents a particularly profound challenge. Technological solutions have historically been limited for those affected by conditions such as locked-in syndrome, Amyotrophic Lateral Sclerosis (ALS), and severe strokes.

However, the startup Cognixion is striving to address this need with an innovative brain monitoring system. This system, coupled with a contemporary interface, promises to streamline and accelerate both communication and interaction.

How the Cognixion One Works

The Cognixion One headset meticulously monitors brain activity. This allows the user to control a cursor, displayed on a visor similar to a heads-up display, in multiple directions.

Selection from menus and options is also achievable without any physical movement. Integrated with modern voice assistants like Alexa, users can efficiently communicate and gain access to information and content commonly enjoyed by others.

Key Features and Capabilities

• Brain Activity Tracking: Precise monitoring of neural signals.
• Cursor Control: Directional control via brain activity alone.
• Hands-Free Operation: Eliminates the need for physical movement.
• Voice Interface Integration: Compatibility with platforms like Alexa.

Understanding the Limitations

Despite its potential, the Cognixion One is not a complete solution. It doesn't represent an instant cure for communication difficulties.

The development of this technology involved a specific journey, and understanding its origins is crucial to appreciating its capabilities and limitations.

Revitalizing Long-Standing Brain-Computer Interface Technology

Individuals experiencing motor impairments present with diverse requirements and abilities, and a wide array of assistive technologies have been developed to address these varying needs. However, a significant portion of these technologies and interfaces are considerably outdated – representing medical equipment that hasn’t benefited from advancements in smartphones and rapid mobile connectivity.

Notably, some of the most antiquated interfaces are utilized by individuals facing the most severe limitations, those whose movements are restricted to head motions, facial expressions, eye movements, or even a single eyelid, as exemplified by Jean-Dominique Bauby, the renowned author of “The Diving Bell and the Butterfly.”

The electroencephalogram, or EEG, stands as one tool within this technological landscape, involving the detection of brain activity through scalp-applied patches that record electrical signals. While valuable in medical and research contexts, EEGs are often characterized by noise and imprecision – proving more effective in identifying active brain areas than pinpointing specific sub-regions of the sensory cortex, for instance. Furthermore, the necessity of wearing an electrode-laden cap, frequently coated with conductive gel, presents a practical barrier to prolonged, everyday use.

Despite these challenges, cognitive function frequently remains intact even among those with profound physical disabilities – a fact substantiated by EEG studies. This observation prompted Andreas Forsland, co-founder and CEO of Cognixion, to explore further potential applications for this established technology, questioning whether an EEG-based brain-computer interface could serve as a functional communication system.

His initial exploration of EEG for assistive purposes occurred within a research study approximately five years prior. The investigation focused on alternative methods for controlling an on-screen cursor, including an accelerometer for head movement detection, with EEG readings integrated as an additional input signal. However, this approach did not yield substantial progress.

A contemporary laboratory setting featuring an EEG cap connected to a receiver and laptop – illustrating a typical EEG application. Image Credits: BSIP/Universal Images Group via Getty Images

Forsland detailed the obstacles encountered: “Existing EEG systems, designed for read-only operation, are inadequate for real-time interfaces; other headsets suffer from slow sampling rates and insufficient accuracy. The most effective BCIs are confined to laboratory settings, utilizing wet electrodes – a messy and impractical solution. The core challenge lies in replicating laboratory-grade performance with dry, passive electrodes, demanding significant engineering innovation.”

Forsland and his team discovered that the limitations were not inherent to the EEG technology itself, but rather to its conventional implementation. Traditional brain monitoring is geared towards diagnosis and research, not real-time feedback. Employing an EEG for this purpose would be akin to using a tractor in a speed contest. EEGs often involve a slow, comprehensive assessment of multiple brain regions, potentially spanning several seconds, and the resulting signal is analyzed using outdated statistical methodologies. Consequently, Cognixion began by re-evaluating both of these practices.

Accelerating the scanning process is more complex than simply increasing sensor speed. Inferring brain activity necessitates the collection of a specific volume of data. Forsland introduced an active element to this passive data collection, employing rhythmic electrical stimulation that is subtly altered by the brain region’s state – functioning similarly to echolocation.

The Cognixion One headset, showcasing its dry EEG terminals. Image Credits: Cognixion

Signals are detected using a customized array of six EEG channels positioned in the visual cortex area (located at the back of the head), and a machine learning model is employed to interpret the incoming data. By running a convolutional neural network locally on an iPhone – a capability only recently achievable – the system can efficiently extract a signal and generate accurate predictions, resulting in faster and more fluid interactions.

This results in sub-second latency with 95-100% accuracy, all within a wireless headset powered by a mobile phone. “The speed, accuracy, and reliability are now reaching commercially viable levels – we are able to match the performance of the leading current EEG systems,” stated Forsland.

Dr. William Goldie, a clinical neurologist with decades of experience utilizing and researching EEGs and other brain monitoring techniques (and a voluntary contributor to Cognixion’s headset development and testing), provided a favorable assessment of the technology.

“There is demonstrable evidence that brainwave activity responds to thought patterns in predictable ways,” he observed. This type of stimulation and response was investigated in prior years. “It was intriguing, but at the time it resided in the realm of speculation. Now, it is being revisited with these specialized techniques and the computational power available today. I believe this area is poised to open up in a way that could have a profoundly positive clinical impact.”

BCI and the User Interface

Forsland’s initial statement – “We are a UI company” – is key. Even groundbreaking advancements in neural interfaces are rendered less impactful without a focus on practical application. Specifically, the goal is to empower individuals with significant motor impairments to communicate efficiently and effortlessly.

Unfortunately, current assistive technologies often fall short. Devices like puff-and-blow tubes and switches require users to painstakingly maneuver a cursor, selecting letters one at a time.

While gaze detection represents an improvement, it isn't universally viable due to potential limitations in eye function. Furthermore, superior eye-tracking systems, such as those found in Tobii Dynavox tablets, often lack portability.

The question arose: why shouldn’t these interfaces offer the same seamless experience as conventional technology? The team dedicated themselves to developing a user interface specifically tailored to the capabilities of their advanced EEG system.

Image Credits: Cognixion

Their approach blends elements of established methods with the functionality of contemporary virtual assistants. A radial design is central, prioritizing rapid responses and addressing frequent user needs. The entire system operates within an iPhone application, with the display mirrored onto a visor, functioning as both a heads-up display and an outward-facing screen.

Simple, frequently used options – affirmations, negations, expressions of gratitude – are readily accessible, requiring minimal cognitive effort. Users can also store pre-defined phrases, such as names or menu requests, for quick retrieval.

A predictive keyboard further streamlines communication, suggesting words and sentences to minimize the need for complete spelling. This allows for faster input of common terms.

Forsland recounted a compelling test case: “We observed individuals who typically require 30 minutes to make just two selections using switches. When we fitted a person with cerebral palsy with the headset, she was able to type her name and initiate playback within two minutes.” He described the emotional impact, stating, “It was profoundly moving; everyone present was in tears.”

Goldie highlighted the adaptive nature of the system. “There’s a period of adjustment involved,” she explained. “The system learns to recognize user patterns, but simultaneously, it also guides the user towards optimal interaction. It’s a reciprocal training process – a continuous feedback loop.”

A Powerful Voice: Utilizing Assistive Technology

Chris Benedict, a DJ, public speaker, and advocate for individuals with disabilities, has discovered significant advantages through the use of innovative technology. He lives with Dyskinetic Cerebral Palsy, which impacts his motor skills and speech. However, this hasn't hindered his ability to pursue his passions.

Despite these challenges, Benedict actively DJs at events and effectively communicates his experiences, even utilizing email correspondence. A demonstration of the technology is available in the accompanying video.

Image Credits: Cognixion

“The headset proves particularly useful, though it isn’t a constant necessity, in facilitating my communication,” Benedict shared. “It’s especially beneficial when a swift response is needed or when I’m in a loud environment – a common occurrence as a DJ.”

He further explained that pairing the headset with a Bluetooth speaker allows him to project his voice clearly, stating, “I can be the loudest person in the room.” He consistently carries a speaker, recognizing the potential need for music at any moment.

Addressing Gaps in Existing Assistive Tech

The advantages provided by this headset highlight shortcomings in current assistive technologies, as well as aspects that are often overlooked.

“This device allows me to communicate while maintaining eye contact with the person I’m addressing, thanks to the visor design,” Benedict noted. “I avoid the disconnect of having to look at a screen positioned between us, which significantly enhances my ability to connect with others.”

The headset’s form factor also offers practical benefits. “Being a headset, it eliminates concerns about navigating doorways or damaging equipment attached to my wheelchair,” he said. “It’s well-balanced, preventing strain on my neck or head.”

The initial setup included a calibration process that measured Benedict’s individual range of motion. This ensured the on-screen keyboard and options were optimally configured for his specific needs. He emphasized the importance of recalibration, as his range of motion can vary daily.

Seamless Integration with Familiar Systems

The Cognixion interface incorporates Alexa, a voice assistant widely adopted by people with disabilities due to its affordability and compatibility with numerous devices. Benedict appreciates this integration, as he already utilizes Alexa for smart home control and other applications.

“Other systems often lack this capability, or if it’s available, the process is overly complex,” he explained. This streamlined integration simplifies his daily life and enhances the overall user experience.

Future Development and Accessibility

As highlighted by Benedict’s experience, the Cognixion device presents a valuable solution for individuals with specific needs. The expectation is that it will gain acceptance as a crucial component within the broad spectrum of available assistive technology.

Forsland emphasized the company’s dedication to collaboration with the user community. This includes direct feedback from users, guidance from clinical advisors such as Goldie, and input from specialists like speech therapists, all aimed at optimizing the One headset’s performance.

A significant challenge, common to many devices in this category, lies in accessibility – both in terms of financial affordability and logistical implementation.

Cognixion is currently pursuing FDA clearance to facilitate insurance coverage for the headset. Its reliance on a smartphone for processing power helps to keep costs lower than devices with integrated components.

In the interim, the company is partnering with clinical and corporate research facilities engaged in neurological and psychological studies. These are environments where traditional, often unwieldy, EEG systems are typically employed.

Funding has been secured, and further investment is being sought to support ongoing hardware development and medical research initiatives. The company has also been successful in obtaining several grants.

While widespread adoption of the Cognixion One headset may still be several years distant – due to the FDA’s rigorous approval process – this timeframe allows for continued refinement and the integration of emerging technologies.

Unlike many assistive devices, such as switches or joysticks, the Cognixion One’s capabilities are primarily constrained by software. Therefore, advancements in algorithms and user interface design will yield substantial improvements.

While anticipation surrounds the development of brain-computer interfaces by companies like Neuralink, Cognixion has already delivered a functional solution for a population that stands to benefit significantly.

Further Information

More details about the Cognixion One headset, and opportunities to stay informed about its progress, can be found on the company’s website.