Abstract
Peripartum depression (PPD) affects 12 to 25% pregnant women worldwide, yet screening often misses real-time symptom changes. Digital phenotyping (DP) offers a promising support, using data like text entries or sleep tracking to detect PPD. This review (PROSPERO: CRD42023461325) evaluated 14 studies, highlighting the substantial potential of personal history and semi-random ecological-momentary data. Future work should focus on improving models and advancing their translation into clinical settings for broader impact.

This publication has 50 references indexed in Scilit: