US Researchers Build AI Tool to Spot PTSD in Children Through Facial Cues

The AI tool analyzes subtle changes in facial expressions. It was trained on over 1,80,000 video frames per child and detects minute muscle shifts linked to emotional distress.

In a significant healthcare-AI breakthrough, researchers at the University of South Florida (USF) have developed an AI-powered facial recognition tool capable of detecting post-traumatic stress disorder (PTSD) in children.

The AI tool analyzes subtle changes in facial expressions. According to the researchers, this is the first study globally to ensure doctor-patient confidentiality while enabling context-specific PTSD classification.

Diagnosing PTSD in children has historically been difficult due to limited communication skills and emotional awareness. “The facial expressions of the children I interviewed visibly intensified during trauma sessions,” said Alison Salloum, professor at USF’s School of Social Work, who led the study with AI expert Shaun Canavan.

Further, Canavan, associate professor at USF’s Bellini College for AI, Cybersecurity, and Computing, created an AI system that prioritizes patient privacy by blurring identity markers and focusing solely on physical data like head pose, gaze, and specific facial movements.

Reportedly, the AI was trained on over 1,80,000 video frames per child, detecting minute muscle shifts linked to emotional distress.

“Data like this is incredibly rare for AI systems, and we’re proud to have conducted such an ethically sound study. That’s crucial when you’re working with vulnerable subjects. Now we have promising potential from this software to give informed, objective insights to the clinician,” Canavan noted.

Additionally, Salloum clarified the system isn’t designed to replace clinicians but could serve as a valuable tool for providing real-time feedback during therapy sessions without repeated, potentially distressing interviews.

The researchers noted that children exhibited more expressive reactions during clinician-led conversations than with their parents, often due to reluctance or shame.

Currently, the team is working to eliminate age, gender, and culture-based biases within the AI system and believes the platform could be adapted to diagnose other mental health conditions, such as depression, anxiety, and ADHD, in children.

Stay tuned for more such updates on Digital Health News