Researchers at the University of Vermont (UVM) have used artificial intelligence (AI) that helps detect hidden depression in young children. Depression, if left untreated, can increase the risk of substance abuse as well as suicide later in life.
A machine-learning algorithm used by the researchers can help detect the potential signs of depression and anxiety in the young children’s speech patterns, offering an easy and prompt way to diagnose conditions that are a bit difficult to recognize and often overlooked in young people.
The new research was published in the Journal of Biomedical and Health Informatics.
At least one in five children suffer from depression and anxiety, which is collectively called “internalizing disorders.”
Young children cannot reliably express their emotional suffering so adults must be able to infer their mental status and recognize potential mental health issues. Failure to recognize the signs can contribute to children missing out professional help.
Clinical psychologist at the UVM Ellen McGinnis said, “We need quick, objective tests to catch kids when they are suffering. The majority of kids under eight are undiagnosed.”
Early diagnosis is important because kids often respond well to treatment while their brains are still developing. However, if left untreated, they get vulnerable to substance abuse and even suicide in later life.
Ellen McGinnis and Ryan McGinnis, study author and biomedical engineer at UVM, have been looking for ways to use AI and machine-learning algorithm to make a diagnosis of depression and anxiety in children faster and more reliable.
The team of researchers used Trier-Social Stress Task, an adapted version of a mood induction task, which causes feelings of anxiety and stress in the subject.
Ellen McGinnis said, “The task is designed to be stressful, and to put them in the mindset that someone was judging them.”
They then used a machine-learning algorithm to analyze and evaluate the statistical features of the audio recordings of children’s story and relate them to their diagnosis.
The researchers found that the algorithm was extremely successful at diagnosing depression and anxiety in young children.
Ryan McGinnis said, “The algorithm was able to identify children with a diagnosis of an internalizing disorder with 80 percent accuracy, and in most cases that compared really well to the accuracy of the parent checklist.”
The algorithm can also give the results much more quickly. It requires just a few seconds of processing time after the task is completed to provide a diagnosis. Ellen McGinnis said the next step is to develop a speech analysis algorithm that can be used as a useful screening tool for clinical purpose, probably by using a smartphone app that could record and analyze results promptly.
