They say that the eyes are the windows to your soul. Although humans can pick up on a variety of emotions based on eye contact, it turns out that artificial intelligence (AI) might be even better at it. A team of scientists has developed an AI that can determine the mental health of cancer patients based strictly on their eye movements, Engadget reports.
The neural network was surprisingly accurate given its experimental nature. Depending on what follow-up test was used to screen the person for anxiety and depression, the AI had an accuracy rate between 93.8 and 95 percent.
That being said, the study only included 25 participants. With such a small sample size, it’s impossible to generalize the results to the entire population. Although more work needs to be done to confirm the findings with a larger group, the early results look promising for this innovative use of AI.
Mental Health Checkup
Those battling cancer are often faced simultaneously with mental health challenges. It’s common for these individuals to deal with issues like anxiety and depression both during their course of treatment and after it. Survivors may suffer from long-term mental health symptoms that often go undiagnosed.
As such, performing regular mental health checkups is an important part of the treatment plan for anyone who’s dealing with cancer. One thing becoming clear is that human healthcare workers won’t always be accessible to perform psychological assessments. Whether it’s because of the overloaded healthcare system or something like the COVID-19 pandemic, patients can (unfortunately) slip through the cracks.
Artificial intelligence, on the other hand, is always accessible. Researchers believe that it can be used as a tool to perform mental health checkups when a human provider isn’t available.
Their approach, which was documented in the journal Translational Psychiatry, relies on multiple algorithms. The neural networks study the eye movements of patients wearing eye-tracking glasses following surgery as they examine pieces of art.
The AI analyzed metrics like gaze and pupil position to determine the likelihood of the person reporting mental health concerns on future screening questionnaires. Several different tests were used to examine topics like hope, anxiety, and mental wellbeing.
More Research Needed
As noted, the AI eye-tracking tech in question was demonstrated in a study made up of just 25 participants. That means the findings aren’t reliable enough to recommend the approach for use with real patients.
However, the results are promising and warrant further consideration. If future studies are able to show that the high accuracy rate is maintained with a larger group, the tech could become a very useful tool for healthcare practitioners.
Engadget’s Jon Fingas notes that there are a variety of potential uses. For instance, it could allow post-surgical patients to recover at home while providers check-in on their mental health remotely to monitor for warning signs.
This is yet another example of how AI has the potential to impact healthcare. Keep an eye on this space in the coming years as more technologies like this one continue to emerge.