News Report | December 31, 2020
Facebook Activity Analysis Can Predict First-Episode Mental Health Hospitalizations Up 18 Months Prior To Diagnosis
A new group of machine learning algorithms analyzing Facebook posts identify users who were subsequently hospitalized for first-episode schizophrenia spectrum disorders (SSD) or mood disorders (MD) up to 18 months after the posts were made. Recruitment for the study took place between March 2016 and December 2018. The most notable differences in posts between healthy volunteers (HV) in the study and those who were hospitalized for SSD or MD were as follows:
- SSD and MD participants were more likely to use swear words in comparison to HV.
- SSD used more perception words – such as “hear . . .