AI Detects PTSD from Social Media with 83% Accuracy in COVID Survivors

Researchers have demonstrated the potential of using social media data and AI to detect post-traumatic stress disorder (PTSD) in COVID-19 survivors. By analyzing millions of tweets, they achieved an 83% accuracy rate in classifying posts as PTSD-positive, based on specific keywords related to trauma symptoms. This study, published in Scientific Reports, highlights the effectiveness of machine learning techniques as an early screening tool for mental health conditions.

Methodology and Findings

The study analyzed 3.96 million tweets from users who mentioned they were COVID-positive between March 2020 and November 2021. The research team employed various machine learning classifiers, including Support Vector Machine (SVM), Naïve Bayes, K-Nearest Neighbor, and Random Forest, to identify posts indicating PTSD symptoms. The SVM classifier achieved the highest accuracy rate of 83.29%.

The researchers focused on identifying key PTSD symptoms such as:

• Re-experiencing: Flashbacks, nightmares, vivid dreams.

• Hyperarousal: Agitation, hypervigilance, irritability.

• Avoidance Behavior: Avoiding reminders of the trauma.

• Other Symptoms: Anxiety, depression, suicidal thoughts, and fatigue.

Tweets were classified as ‘PTSD Positive’ if they contained both COVID-19 mentions and PTSD-related keywords. Posts that mentioned PTSD symptoms unrelated to COVID-19 were categorized as ‘PTSD Negative.’

Implications for Mental Health Intervention

The study underscores the significant mental health impact of COVID-19 and the potential for social media platforms to serve as early screening tools. Professor Mark Lee, co-author of the study from the University of Birmingham, emphasized the importance of early detection: “Our findings demonstrate that social media data can provide a valuable means of identifying people who are at risk of PTSD – enabling early screening and prompt medical action.”

The researchers believe that with further refinement, these machine learning techniques could be applied to detect other health conditions, broadening the scope of social media as a tool for public health monitoring.

Understanding PTSD

PTSD is an anxiety disorder that can develop after experiencing or witnessing a traumatic event, such as a natural disaster, car accident, or assault. Common symptoms include flashbacks, nightmares, and severe anxiety. The World Health Organization (WHO) and the American Psychiatric Association (APA) both recognize PTSD as a legitimate mental health condition requiring medical attention.

Future Research and Applications

Co-author Dr. Mubashir Ali highlighted the potential for expanding this research to other health issues: “Our analysis indicates that the pandemic took its toll on people’s mental health, flagging the possible impact of symptoms such as anxiety, insomnia, and nightmares rampant among COVID-19 survivors.” With continued research, similar techniques could be employed to monitor other conditions, providing valuable insights for early intervention.

This study demonstrates the promise of integrating AI and machine learning into mental health care, potentially transforming how we approach the detection and treatment of psychological disorders in a post-pandemic world.