THE NEURO-WHAT?

Basically, neurocomputing is the study of the internal functioning of the human brain and its implementation in artificial systems for emulation. In a way, we are talking about generating a "digital twin" of the human brain, but a human brain that exists within a computer system where we can digitally simulate its operation. While this may sound like science fiction, there are already hundreds of researchers in the Basque Country focused on this task. Additionally, projects like the European Human Brain Project, which has been producing positive results for a decade, and Basque projects continue to impress with their successes.

But why would we want to emulate a human brain on a computer? From a computer engineering perspective and within the development of new artificial intelligence, the answer can be found in the insights provided by last year's Princess of Asturias laureates (Geoffrey Hinton, Yann LeCun, Yoshua Bengio, and Demis Hassabis). Thanks to their research on how the brain works and efforts to replicate it in mathematical models executable on current computers, particularly in understanding our memory and visual cortex, there have been advances in the last five years that were unimaginable just decades ago: in applications for voice recognition, computer vision, and natural language processing, among others. Who hasn't used ChatGPT for professional purposes? Without neurocomputing, this progress would not have been possible.

TECHNOLOGICAL ADVANCES BEYOND CHATGPT

However, what do these advances have to do with disability or rehabilitation? The answer is evident: like other use cases, these advances in artificial intelligence have also been a turning point in clinical, diagnostic, therapeutic, and especially in improving the lives of people with disabilities or special needs.

Focusing on each of these aspects, a couple of years ago, Ibermática participated in the DeepBrain project with Deusto Sistemas and BioAraba, where a digital brain in a 3D environment was created. Based on brain scans, the system identified whether there was a risk of brain damage or not. Furthermore, thanks to a predictive AI system, it is possible to forecast the progression of such pathology in the future, enabling it to be treated in the best possible way. To achieve this, with the assistance of neuroscientists, data scientists, as well as graphic designers, research was conducted on how the most advanced automatic AI systems could distinguish different brain areas, automatically segment them, represent them in a 3-dimensional format, and search for patients with similar clinical episodes in the past, analyzing their evolution.

Another clinical project led by Ibermática was the SUPREME project, where, together with Tecnalia, a tool capable of identifying early signs of Alzheimer's and predicting the degree of disease progression was developed. These are systems that take brain images and can generate 3D maps of the brain, analyze them, and identify anomalies or pathologies in these digital twins; in other words, they directly impact clinical diagnosis.

THE CHARLES XAVIER OF OUR TIME

If we focus on the therapeutic aspect, there’s another technology associated with neurocomputing with multiple practical applications, especially in the video game market and entertainment industry, known as the famous EEG headsets, also called Brain- Computer Interfaces (BCI). These systems are capable of reading the brain’s electrical signals (if you’ve seen at least one season of Grey’s Anatomy, you’ll know what an electroencephalogram is) and interpreting them. In fact, there are already systems where, to control characters on consoles, robots, machines, equipment, or even drones, you don’t need to use controllers or keyboards; you just have to think about what you want to happen.

But this technology goes much further. Think about the revolution of being able to directly interpret our brain signals: our thoughts, preferences, intentions, and commands. It won’t be necessary to speak to our beloved friends Alexa or Siri using voice anymore; instead, we’ll be able to interact directly with our devices using our thoughts.

In this line, the Ibermática Innovation Institute, in collaboration with the University of Deusto and Virtualware, built a system that precisely allowed understanding the motor intentions (moving right, left, up, or picking up an object…) of users to control robots in a virtual environment. This same industrial case is already being applied for the recovery of motor or mental problems in rehabilitation settings or in situations where patients need to relearn or reinforce certain tasks by retraining their brains.

In fact, BCI systems allow these patients to think about limb movement, imagine mental scenarios, engage in exercises for recovering brain plasticity, or relearn speech. With the help of a virtual environment where their thoughts are transformed into actions, sounds, or text directly, rehabilitation accelerates exponentially, and feedback is obtained from patients without them needing to verbalize it. Therefore, extracting brain electrical signals, understanding them, and modeling them using mathematical algorithms that current systems can understand is the main objective of neurocomputing. The most impactful application of this technology is the direct connections of microchips connected to certain human neuronal circuits (i.e., the connection of artificial neural networks directly with human neural networks), which is enabling people with motor disabilities to walk, pick up, and manipulate objects. It is even being applied to people with certain levels of visual impairment so they can see.

Furthermore, the latest research in neurocomputing allows for the preventive identification of pathologies such as cognitive and affective disorders, evaluation of brain damage in neurodegenerative diseases or traumas, epilepsy, tumors, or paralysis. Thanks to the interaction between intelligent systems, mathematical models, AI, and humans, these technologies can mitigate and aid in the rehabilitation of these problems.

COMING SOON TO YOUR NEAREST NEWSSTAND

Hold onto your hats, folks! Neurocomputing is breaking out of the lab and weaving its way into our daily lives. Companies like BitBrain, Neurable, OpenBCI, and EMOTIV are rolling out more affordable and powerful EEG headsets. Big players like Elon Musk's Neuralink are even pushing brain chips! Apple, true to its innovative spirit, just patented an EEG device that pairs up with your regular headphones. In the research community, there are tons of software platforms to tinker with these headsets and whip up cool applications. And get this, we're seeing stories every day about people overcoming major communication hurdles thanks to this brainy tech.

According to the wise words of Yuval Noah Harari in his book "Sapiens: A Brief History of Humankind": "Out of all the projects currently in the works, the most revolutionary is the attempt to design a two-way brain-computer interface that lets computers read the electrical signals of a human brain and simultaneously transmit signals that the brain can read."

Now, here's the scoop on the current headache with existing systems (especially those non-invasive brainwave-reading headsets): they struggle with making brain signals universally understandable for today's AI. It's like each system has to become buddies with your unique "thought style" – no universal language. Flashback 15 years, when voice recognition or speech-to-text systems needed a personal touch for each user, and the same goes for EEG systems today. But hey, look at us now – we can dictate anything into our phones, and they transcribe it like champs!

WHAT LIES AHEAD (IN THE NEAR FUTURE)?

Here's the kicker – there's a new tech in town that could solve the universalization problem. We're diving into a technology that can analyze the nitty-gritty details of each person's brain signals and configure them into a universal "wave": quantum computing. In fact, there's a buzz about whether our brains might work a bit like quantum computers. If that's the case, slapping quantum computing into the mix could send neuroscience on a wild ride, bringing AI even closer to its holy grail: General AI.

We're betting the farm that the combo of neurocomputing, artificial intelligence, and quantum computing is going to shake up the future of health tech, diagnostics, therapies, and advanced rehab solutions.

It's not just wishful thinking – it's happening. Here in Biscay, the brainiacs at Ayesa-Ibermática, along with Lantik, are teaming up neuroscientists, engineers, and the arts to dive deep into this tech. Wanna join the revolution?