Illustration by Angelika Manhart
Standard Deviations

A new way to treat mental illness

By Z. Sage Chen

Millions of Americans suffer from mental illness—depression, schizophrenia, autism or similar disorders. The global cost of mental health, according to a Lancet report, is estimated to be $16 trillion by 2030. A new way to tackle this global crisis has emerged — and it sounds like the stuff of science fiction.

Mental illness, according to neuroscience research, has been linked to dysfunctional brain circuits that control mood, memories, and attention. Many scientists believe that if you can “read the mind” of a person with mental illness—see how they respond to memory tasks, how they react to stress in sleep, or how they interact with others—you might be able to diagnose the disorders more precisely. You might even intervene in their brains to redirect their mental circuitry and keep their illness at bay.

The mind consists of manifestations of thought, perception, memory and imagination that take place in different areas of the brain. Multiple parts of the brain work together to drive the thought process. Mental illness may arise as a result of dysfunction of one or multiple circuits, causing the path of thoughts to go awry. Advances in artificial intelligence (AI)-powered neuroscience suggest that mind-reading may be scientifically possible, and the associated technologies are already being developed, fast.

Neurotechnology for mental health applications is no longer a question of if, but when. Early last year, Elon Musk’s new company, Neuralink, announced the blueprint for a small device that fits outside the human head and wirelessly receives information from small and flexible electrode threads implanted in the brain, with as many as 3,072 electrodes per array. These so-called “invasive” neural interfaces can potentially help paralyzed people use phones or computers.

Several other tech companies, including Google and Facebook, have already invested in brain or mind-related research in the past few years. CTRL-Labs, an AI company recently acquired by Facebook, is dedicated to the development of noninvasive mind-reading devices, as simple as a wristband, to mentally control computers or prosthetics. Kernel, a human intelligence company founded by tech pioneer Bryan Johnson, develops new-generation neuroprosthetics for patients suffering from strokes, Alzheimer’s or concussions.

Brain-machine interfaces that can read and influence brain activity have been proposed — and used — for nearly a century. German psychiatrist Hans Berger first used an electroencephalograph (or “EEG”) to study electrical activity in the human brain in 1924. These “brain waves” can be used to determine a person’s attentional or sleep state. Later technologies allowed scientists to collect recordings directly from neurons by inserting electrodes directly into the brain.

Today, brain-machine interface technologies—invasive (like the electrodes) or noninvasive (like the EEG)—have various applications ranging from motor control, seizure suppression, anesthetic drug delivery, and pain management to mood restoration.

Brain-machine interfaces could also help neuropsychiatric disorders by restoring the damaged or lost brain and emotional functions that are fundamental to a healthy mind. Take memory, arguably the most fundamental cognitive function. Memory impairments are common in many mental disorders, like dementia and post-traumatic stress disorder. Whether it is remembering the path from our home to work, or recollecting an unfamiliar face at a party, a healthy brain accomplishes the task with ease, relying on a seahorse-shaped structure called hippocampus. The hippocampus, located at the temporal lobe of the cerebral cortex, is known as the brain’s GPS or “cognitive map” that guides us in daily navigation, planning and learning.

How do we memorize things and recall them? Every episodic memory creates a neural “fingerprint” about the “where” and “what” of information in the brain. During memory recall, similar neural patterns appear in the brain scan. The brain of a schizophrenic or autistic patient, however, may respond to the same memory task differently from the normal brain. The differences will be picked up by studying the patterns of electrical activity of individual neurons, called neural codes. As in a crime lab, memory reading attempts to reconstruct the trace of neural fingerprints from these neural codes.

Cracking neural codes, therefore, is key to the success of mind reading. We’re already making important advances. In my laboratory at New York University, for instance, we collaborated with Professor György Buzsáki, an expert specializing in brain rhythms and episodic memories, to develop techniques for reading information encoded in the rodent hippocampus. We demonstrated that we could reconstruct a rodent’s spatial memories by eavesdropping on the animal’s brain activity in sleep. We could also predict the animal’s future decision-making in maze. Our new memory-reading techniques can process information from hundreds of thousands of wires implanted in the hippocampus in real time. This technology can potentially detect memory patterns in the human hippocampus or cortex, and distinguish the differences between a healthy and a pathological brain.

Neurotechnologies like the ones we’re developing are moving fast. Scientists and physicians are beginning to test these devices on the diagnosis and potential therapy of mental illness. According to the U.S. Food and Drug Administration, neurostimulation-based therapeutic treatments are on the rise, such as transcranial magnetic stimulation (TMS) for depression and responsive neurostimulation (RNS) for epilepsy. And while these mind-reading technologies may have a profound positive impact on mental health, this uncharted reality invites another question: does mind-reading impose a threat to humanity?

We need to think about the moral and legal implications of mind-reading technologies now. Is memory augmentation ethical for taking exams? Should it be considered cheating in the classroom the same way doping is in sports? Should rewiring brain circuitry to erase memories be regulated by law? As the ethical concerns regarding the use of these technologies are continually debated, an international standard should be established across the globe and reviewed annually to accommodate the emerging issues.

Nevertheless, the neural revolution for mental illness is here — and I’m optimistic. When science and managed technology advance hand in hand, the future for mental health treatment is bright.


Z. Sage Chen is an Associate Professor of Psychiatry, Neuroscience and Physiology in the School of Medicine at New York University, and an investigator at the NYU Neuroscience Institute. He studies memories and sleep at work, while enjoying dancing Argentine tango at night.