BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Los_Angeles
X-LIC-LOCATION:America/Los_Angeles
BEGIN:DAYLIGHT
TZOFFSETFROM:-0800
TZOFFSETTO:-0700
TZNAME:PDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0700
TZOFFSETTO:-0800
TZNAME:PST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20261012T210915Z
LOCATION:Summit Pavilion
DTSTART;TZID=America/Los_Angeles:20261021T173000
DTEND;TZID=America/Los_Angeles:20261021T183000
UID:HFESAM_ASPIRE 2026_sess257_POST263@linklings.com
SUMMARY:Machine Learning for Mental Health Episode Prediction via Wearable
  Devices
DESCRIPTION:Karim Zahed (American University of Beirut)\n\nNearly 970 mill
 ion people worldwide live with mental health (MH) disorders, many involvin
 g recurrent episodes that may be mitigated through timely prediction. This
  review evaluates machine learning (ML) approaches using wearable physiolo
 gical data to forecast MH episodes and examines whether device grade influ
 ences predictive accuracy. Empirical English-language studies were identif
 ied through EBSCO and screened in Rayyan.ai by two independent reviewers, 
 with a third reviewer resolving conflicts. Of 332 records identified, 32 s
 tudies met the inclusion criteria. Most studies focused on depression or s
 tress, and 30 applied AI/ML methods. Findings show that predictive perform
 ance was driven more by ML approach than by device grade. Reported ranges 
 reached 76–87% for depression, 85–98% for stress, 75.6–92.8% for panic-att
 ack risk, and 97–99% for anxiety. Key predictors included heart-rate varia
 bility, sleep, activity, circadian stability, electrodermal activity, and 
 oxygen saturation variability. Multimodal and personalized temporal models
  performed best.\n\n
END:VEVENT
END:VCALENDAR