A mental health assessment usually begins with words.
A doctor or therapist asks how a person has been feeling, how long symptoms have lasted and whether thoughts of self-harm or suicide are present, but a University of Southern California research team is investigating whether the body may provide additional clues that are harder to consciously control.
The researchers are combining information from the brain, the eyes and the skin with artificial intelligence. Their goal is to see whether these biological responses can help distinguish people without depression from those with depression and those experiencing suicidal thoughts.
The work comes from a USC project called PRECOG, short for a project examining neural and behavioral signals linked to responses that may occur before conscious reporting. It started in 2023 and includes researchers from engineering, computing, psychology, linguistics, brain science and psychiatry.
There is a practical reason for taking this approach. Mental health conditions do not usually come with a simple laboratory result, so clinicians depend heavily on what people report, what they observe during an interview and how symptoms fit established diagnostic criteria.
Those methods remain central to good care, but they have limitations. A person may have difficulty describing feelings, may minimize symptoms because of fear or stigma, or may experience changes that are not captured during a short appointment.
The USC researchers designed an experiment in which participants responded to emotionally loaded statements about themselves. They read 160 sentences with different emotional meanings and chose whether they agreed or disagreed, while several instruments quietly recorded what was happening in their bodies.
One instrument measured brain activity with EEG. Small sensors on the scalp recorded rapid electrical changes, allowing the researchers to examine what happened in the fractions of a second after a participant saw emotionally important language.
The analyses found differences in the timing and strength of brain responses among healthy, depressed and suicidal groups. In one study, particularly useful differences appeared roughly 300 to 600 milliseconds after a word was presented, when the brain is interpreting meaning and emotional content.
Another analysis examined specific waves in the EEG recording. Some responses to negative material were weaker in depression, while another signal appeared to change with the severity of suicidal thinking, giving researchers a possible way to separate general depression from additional suicide-related risk.
At the same time, cameras tracked where participants looked on the screen. Eye movements are useful to researchers because people make many of them automatically, and patterns of attention may change when someone is processing negative or personally meaningful information.
The eye-tracking work found that side-to-side gaze patterns were particularly informative. Healthy participants often looked at negative statements in a more organized way, while people with suicidal thoughts tended to show more scattered or less engaged viewing patterns in the experiment.
The researchers also measured skin conductance, which reflects very small changes in sweating. Sweat glands respond to the body’s automatic stress and arousal system, so these changes can provide a physical record of how strongly a person reacts even when the reaction is not obvious from the outside.
Again, negative information seemed especially revealing. Participants with depression and suicidal thoughts showed altered responses to negative words, and computer models used the timing and strength of these reactions to search for patterns associated with the different groups.
The project has led to multiple publications rather than one single study. The eye-tracking research was published in npj Digital Medicine; EEG findings appeared in the Journal of Affective Disorders, Communications Biology and Translational Psychiatry; and work using skin conductance was posted on the arXiv preprint server, where research can appear before formal peer review.
Together, these papers suggest that mental health may leave subtle traces in several automatic biological systems. That is important because a future assessment combining different signals could potentially provide a fuller picture than any one measurement by itself.
Still, the results should not be interpreted as a machine that can read a person’s mind or reliably forecast a suicide attempt. The researchers identified group-level patterns and built classification models, which is very different from proving that a test can predict an individual’s future behavior in real clinical settings.
Specialized EEG and eye-tracking equipment may also be difficult to use in ordinary clinics, and results from AI systems depend strongly on the people and data used to train them. Larger studies would need to include people from different ages, backgrounds and health situations to make sure the models do not work well for one group but poorly for another.
There is also a strong reason to keep human judgment at the center of care. Suicide risk can change quickly and is shaped by many factors that sensors cannot fully capture, including a person’s circumstances, relationships, history, access to support and current level of distress.
For these reasons, the USC team describes the technology as a possible supplement to existing assessment rather than a replacement. If future research confirms the findings, biological signals could eventually help clinicians notice patterns that deserve closer attention and provide another source of information when deciding what support a patient needs.
Overall, the research is innovative because it combines several kinds of automatic body responses instead of searching for one perfect depression marker. The evidence is encouraging but early, and the strongest conclusion today is that the brain, eyes and skin may contain useful clues that deserve further testing—not that scientists have created a definitive test for depression or suicide risk.
If you care about mental health, please read studies about vegetarianism linked to higher risk of depression, and Vitamin D could help reduce depression symptoms.
For more health information, please see recent studies that ultra-processed foods may make you feel depressed, and these antioxidants could help reduce the risk of dementia.
Source: University of Southern California.
