Deciphering Thought: The Silent Revolution of Light

No electrodes, bulky helmets, or humming MRI scanners—just laser light and the rhythm of a human voice reading sentences aloud. Some words are understood by the listener. Others sound like meaningless sounds. And yet, incredible as it may sound, light can distinguish one from the other.

This beam—and the artificial intelligence that deciphers its reflections—has become the basis for a method that could change the way we study brain function. In a paper published in the summer of 2025 in the Journal of Biomedical Optics (Volume 30, Issue 6, DOI 10.1117/1.JBO.30.6.067001), Natalia Segal and her colleagues from the Faculty of Engineering at Bar-Ilan University described a system that can remotely and without contact distinguish how the human cortex responds to meaningful and meaningless speech.

Підписуйтеcь на наш Telegram-канал Lenta.UA - ЄДИНІ незалежні новини про події в Україні та світі

Using a simple setup—a single laser, a single high-speed camera, and a single neural network—the team achieved accuracy comparable to classical neuroimaging methods, but without touching the patient.

A Light Touch to the Mind

Conventional brain imaging methods require bulky equipment like MRI scanners or dense EEG helmets.

Segal's approach is based on a principle familiar to optical engineers but largely unknown in neuroscience: speckle patterns—grainy interference patterns that occur when a laser reflects off an uneven surface.

When light touches human skin, these tiny fluctuations reflect subtle vibrations caused by blood flow and neuronal activity. By training a neural network to analyze thousands of such video recordings, the researchers were able to determine when Wernicke's area, responsible for understanding speech, responds to familiar language and when it responds to unfamiliar language.

The accuracy achieved an AUC of 0.94—an impressive result, indicating that the system truly distinguishes between the two brain states.

"The idea was to replace wires and electrodes with light," says Segal. "We were looking for a way to 'read' the brain safely and remotely. I spent a lot of time in a neurological hospital with my grandmother and noticed how difficult it was for patients to tolerate touch. This solution eliminates the touch problem."

From Physics to Humanity

Natalia Segal's path to this discovery spanned engineering, entrepreneurship, and biomedicine. Before neurophotonics, she built a career combining artificial intelligence and invention: she has received US patents, many of which have been cited by Apple, Samsung, and Microsoft, and one of which was acquired by Google.

Her ideas cover a wide range of technological areas—from proximity authentication systems in wearables to GPU virtualization and privacy indicators for smart glasses.

Segal's technical ingenuity is always balanced by a focus on people and a desire to make technology more humane.

At Fairtility, she led the AI ​​team, whose algorithms now help IVF clinics around the world—work that became the foundation of the FDA-approved CHLOE Blast platform.

Earlier, she founded Sphoonx, an educational startup: its apps for children with ADHD have been downloaded over 100,000 times, and many are preinstalled on school tablets.

At Bar-Ilan University, where she completed her master's degree and is continuing her doctoral research under the supervision of Professor Ze'ev Zalevsky, Segal channeled the same creative impulse into brain research.

Zalevsky is one of Israel's most renowned optical scientists, a member of SPIE and IEEE. His team made neuroscience fluid, free, and more relevant to everyday practice.

Seeing Understanding

The experiments were simultaneously simple and elegant. Volunteers were played short excerpts—in English and in Swedish, which no one spoke—while a camera recorded speckle patterns from a laser spot aimed at Wernicke's area.

Across seven participants, the average recognition accuracy was 89%. Even when tested on new subjects, the model maintained high accuracy, proving that the system truly captures the cortex's response, not just superficial skin vibrations. Control experiments, where the beam was aimed at the forehead, yielded random results—further confirmation that the signal originates from the brain.

In practice, this technology could allow doctors to assess brain function without touching it—from diagnosing comas to restoring speech after stroke.

Someday, a beam of light no stronger than a laser pointer could detect attention span or early signs of neurodegenerative changes—without a single wire.

Beyond the Lab

The power of this achievement lies not only in the data, but also in the idea behind it.

Segal speaks of "lowering the barrier to entry" in neuroscience—of a future where remote cortex monitoring becomes as natural as measuring one's heart rate.

This approach could make brain research widespread—in schools, clinics, even at home.

Her open-source project, SpecklesAI, on GitHub is a rare example of transparency in neuroscience.

Beyond the Lab

The power of this achievement lies not only in the data, but also in the idea behind it.

Segal speaks of "lowering the barrier to entry" in neuroscience—of a future where remote cortex monitoring will be as natural as measuring one's heart rate.

This approach has the potential to make brain research widespread—in schools, clinics, even at home.

Her open-source project, SpecklesAI, on GitHub is a rare example of transparency in such a competitive field. "We wanted others to build on this," says Segal. "Science moves faster when it's shared."

The prospects extend far beyond medicine.

Photonic brain-computer interfaces, recently considered science fiction, are becoming reality.

If thoughts can be deciphered not through electrodes, but through light, it could enable communication with paralyzed patients or create a form of "silent" dialogue between humans and machines.

An inventor at the edge of possibility

Natalia Sigal's work bridges disciplines with rare ease.

In industry, her patents describe the evolution of smart wearables and contextual technologies; in academia, her papers reflect the growth of optical methods driven by artificial intelligence.

And everywhere, the same theme runs through: making technology more human.

"Natalia Sigal possesses a rare gift—combining the precision of AI model architecture with the scale of ideas and an understanding of people," says her colleague.

"She explores the world with the curiosity of a pioneer, with the persistence and analytical acumen of a scientist who doesn't stop until her vision becomes reality."

It is this combination—precision and curiosity, vision and tenacity—that defines her scientific signature.

Her next step is even bolder: an attempt to decode inner speech—not what we say, but what we say in our minds. If this is successful, humanity will gain a new way to communicate with those who have lost their voice due to stroke, paralysis, or Parkinson's disease: silence can be translated into data and heard.

Toward a Bright Mind

Despite its technical complexity, the essence of the experiment remains almost poetic: light reflects meaning. Each changing speckle bears a subtle imprint of understanding, recognition, thought. Capturing this trace required not only engineering talent but also a quiet confidence that the invisible can be made visible. The brain is never silent—one only needs to learn to listen to its light.

References

Segal N., Kalyuzhner Z., Agdarov S., Beiderman Y., Beiderman Y., and Zalevsky Z. (2025).

AI-powered remote monitoring of brain responses to clear and incomprehensible speech via speckle pattern analysis. Journal of Biomedical Optics, 30(6), 067001. https://doi.org/10.1117/1.JBO.30.6.067001