Early Detection of Mental Health Problems

A portion of psychiatric patients relapse when the treatment runs its course, or sometimes before. The psychiatric or psychological evaluation of patients is complicated by patients’ frequent refusal to cooperate with medical specialists. Such clinical evaluations may also be unable to detect subtle changes in the patients’ conditions or may not be timely enough. This disclosure describes techniques, implemented with appropriate permissions, for automatic monitoring of patients to detect changes in the patient behavior that predict a relapse or worsening of their mental health treatment. Various sensors in user devices such as smartphones, smartwatches, or other wearables, as well as other devices such as smart speakers, home automation devices, etc. obtain sound, image, and body-parametric data of the patient and in their environment. The data is analyzed to test for changes in the patient’s behavioral patterns over time (which can be a reflection of changes in the patient’s health); to test if the patient’s visual and aural fields have a basis in reality; to provide feedback to the patient; etc.