Do you know Artificial Intelligence (AI) can now detect anxiety and depression in children?

The simulation of various human intelligence processes by machines and computer systems is referred to as artificial intelligence (AI). This process involves various learning, reasoning and self-correction by the machines that are programmed to think like humans and mimic their actions. With the further development of AI, various sectors like education, pharmacy, engineering, robotics and others are getting positively impacted. And various other researches are being carried out to bring innovative discoveries.

As per the latest research results of the Journal of Biomedical and Health Informatics, a machine learning algorithm or AI can detect signs of depression and anxiety through the speech patterns of young children. I can provide a fast and easy way of diagnosis, which is often difficult to pinpoint and gets overlooked. Reports state that almost 1 in 5 children undergo anxiety and depression, popularly known as internalising disorders. As children under the age of eight cannot explain their emotional suffering, adults need to be more cautious to understand their mental problems. Long queues for appointments with psychologists, incapability of detecting the problem, etc. make the child miss out the needed treatment. This issue raised urgency amidst the researchers and Ellen McGinnis, a clinical psychologist at the University of Vermont in the US came up with a solution. Researchers have been trying long to use AI to get reliable results after diagnosis and finally with an adapted version of a mood induction task called the Trier-Social Stress Task, they have come close to detecting internalising disorders by AI

Test Procedure : A group of 71 children between the ages of three and eight were selected, who were asked to extemporize a 3-minute story and were informed that they would be judged based on its interest quotient. The researchers acted as judges and gave neutral or negative feedback after every speech. Also, a buzzer was fixed after 90 seconds and 30 seconds to remind the children about the time left. The entire ambience was created to induce a stressful situation. The children were also analysed of using a structured clinical interview and parent questionnaire, which helps in identifying internalising disorders in children and are established ways. After analysing statistical features of the audio recordings of each kid's story and examining the child's diagnosis report with the help of AI and machine learning technology, the researchers found that the process could pinpoint the children. Most importantly, the period between the two buzzers was the most predictive ones.

Results and Future Thoughts : The diagnosis was almost 80% accurate and also matched with the parent checklist. The algorithm is swift and takes a few seconds to give the diagnosis. McGinnis, who was among the researchers explained that the next step for them will be to produce the speech analysis algorithm as a universal screening tool for clinical use, maybe through a smart-phone app that can record as well as present results immediately. The researchers are also planning to combine it with motion analysis or other diagnostic tools to better detect children at risk even before the parents can suspect anything.

Conclusion

Meanwhile, AITS has recently set up a machine learning (ML) Lab and the institute constantly keeps on arranging seminars, development programs for the students and the teachers on AI and ML. This latest development in the AI field will surely create a positive impact on AITS’ teaching methods creating a healthy ground for the students to grow.