

![]() |
|||||||||||||
|
| All | Since 2020 | |
| Citation | 6651 | 4087 |
| h-index | 26 | 21 |
| i10-index | 174 | 83 |
Search
News & Updation
AI-POWERED CONVERSATIONAL AGENTS FOR PERINATAL MENTAL HEALTH SCREENING AND INTERVENTION: A NARRATIVE REVIEW
Manju H, Kalaivanan S., Venkateswaramurthy N.*
ABSTRACT Perinatal depression and anxiety are common, clinically consequential, and frequently underdetected despite recommendations for repeated assessment during pregnancy and postpartum. Artificial intelligence (AI)-powered conversational agents may reduce access barriers through private, asynchronous dialogue for standardized screening, psychoeducation, brief psychological support, monitoring, and referral navigation. This narrative review evaluates their clinical role, evidence base, architectures, safety requirements, and implementation priorities. Peer-reviewed studies, accepted manuscripts, explicitly labeled preprints, and authoritative guidance available through 24 July 2026 were considered. Traceable reports describe Dr Joy, Wysa, Rosie, Juno, Moment for Parents, REA, a postpartum adaptation of Woebot, and early large language model prototypes. The literature indicatesgenerally favorable usability and acceptability, but most perinatal studies are developmental, observational, qualitative, or pilot evaluations; durable clinical effectiveness remains uncertain. Current evidence favors adjunctive use within clinically governed stepped-care pathways rather than autonomous diagnosis or treatment. Validated instruments can be delivered conversationally only when wording and scoring are preserved and the administration method is appropriately validated. Risk results must connect to transparent escalation, clinician review, and emergency resources. Generative models add risks of hallucination, inconsistency, privacy loss, bias, anthropomorphic overtrust, and model drift. Future studies require adequately powered pragmatic trials, prespecified safety outcomes, diverse multilingual samples, longer postpartum follow-up, economic evaluation, and standardized reporting. Conversational AI could extend perinatal mental health services if clinical governance, equity, interoperability, and human accountability are built into the intervention. Keywords: Artificial intelligence; conversational agent; chatbot; perinatal mental health; postpartum depression; screening; digital intervention. [Download Article] [Download Certifiate] |
