Self-administered interventions based on natural language processing models for reducing depressive and anxious symptoms: Systematic review and meta-analysis (Preprint) DOI Creative Commons
David Villarreal‐Zegarra, Jackeline García-Serna, Gleni Quispe-Callo

et al.

Published: April 15, 2024

BACKGROUND The introduction of Natural Language Processing (NLP) technologies has significantly enhanced the potential self-directed interventions for treating anxiety and depression by improving human-computer interactions. Despite these advancements, particularly in AI Large Models (LLMs), robust evidence validating their effectiveness remains sparse. OBJECTIVE To determine whether based on NLP models can reduce depressive symptoms. METHODS Our study was a systematic review, protocol registered PROSPERO (CRD42023472120). databases we used review are Web Science, SCOPUS, MEDLINE (via PubMed), PsycINFO EBSCO), IEEE Xplore, EMBASE Cochrane Library. quality included studies assessed using JBI Critical Appraisal Tools. RESULTS 21 articles were selected 16 meta-analysis each outcome. overall showed that self-administered more effective reducing symptoms (SMD=0.819; 95%CI: 0.389-1.250; p<0.001) (SMD=0.272; 95% CI: 0.116-0.428; p=0.001) compared with various control conditions. In subgroup analysis, AI-based shown to be (SMD=1.059 [0.520 1.597]; (SMD=0.302 [0.073 0.532]; p=0.010) pooled Also, NLP-based outperform psychoeducation bibliotherapy both (SMD=1.481 [0.368 2.594]; p=0.009) (SMD=0.561 [0.195 0.927]; p=0.003). addition, than waitlist or no intervention anxious (SMD=0.196 [0.042 0.351]; p=0.013). CONCLUSIONS findings support usefulness self-applied alleviating widely prevalent mental health problems such as CLINICALTRIAL Protocol (CRD42023472120)

Language: Английский

Evaluating the Quality of Psychotherapy Conversational Agents: Framework Development and Cross-Sectional Study (Preprint) DOI Creative Commons
Kunmi Sobowale,

D. Humphrey

JMIR Formative Research, Journal Year: 2024, Volume and Issue: unknown

Published: Aug. 20, 2024

Language: Английский

Citations

0
Behavior Change Support Systems for Self-Treating Procrastination: Systematic Search in App Stores and Analysis of Motivational Design Archetypes (Preprint) DOI
Jeanine Krath, Manuel Schmidt-Kraepelin,

Katharina Schmähl

et al.

Published: Aug. 8, 2024

BACKGROUND The phenomenon of procrastination refers to an individual’s conscious decision postpone the completion tasks despite being aware its adverse consequences in future. Extant research this field shows that is associated with increased levels anxiety and stress likelihood developing depression calls for development suitable interventions support individuals making lasting positive changes their behaviors. In parallel, practice has produced a plethora behavior change systems (BCSSs) aim provide low-threshold, accessible alternative in-person therapeutic approaches. Most these BCSSs can be considered motivational combine functional, utilitarian components hedonic eudaimonic design elements empower self-treatment. Although early studies have suggested potential benefits such BCSSs, on understanding specific characteristics self-treating still infancy. OBJECTIVE response gap between research, we aimed analyze systemize multitude practical efforts self-treatment identify main archetypes emerged. METHODS We conducted 3-step approach. First, identified 127 apps through systematic screening process German US Apple App Store Google Play Store. Second, systematically coded terms techniques targeted by functional or elements. Third, 2-step cluster analysis combat procrastination. RESULTS A variety designs been developed implemented practice, our five archetypes: (1) structured progress monitor, (2) self-improvement guide, (3) productivity adventure, (4) emotional wellness coach, (5) social focus companion. target different psychological determinants successfully use extend beyond current state research. CONCLUSIONS results study foundation future endeavors examine comparative effects develop more effective tailored individual needs. For practitioners, findings reveal contemporary space may serve as blueprints guide systems. seeking health professionals treating procrastination, systemizes landscape apps, thereby facilitating selection one best aligns patient’s

Language: Английский

Citations

0
MULTIWD: Multiple Wellness Dimensions in Social Media Posts DOI Creative Commons
MSVPJ Sathvik,

Muskan Garg

Published: May 17, 2023

<p>The concept of wellness, as proposed by Halbert L. Dunn, recognizes the importance multiple dimensions, such social and mental well-being, in maintaining overall health. Neglecting these dimensions can have long-term negative consequences on an individual's well-being. In context traditional in-person therapy sessions, efforts are made to manually identify underlying factors that contribute disturbances, factors, if triggered, potentially lead severe health disorders. Our research focuses introducing a meticulous task aimed at identifying indicators wellness detecting their presence self-narrated human writings Reddit media platform. We mentioned Ethics Broader Impact.</p>

Language: Английский

Citations

1
MULTIWD: Multiple Wellness Dimensions in Social Media Posts DOI Creative Commons
MSVPJ Sathvik,

Muskan Garg

Published: May 17, 2023

<p>The concept of wellness, as proposed by Halbert L. Dunn, recognizes the importance multiple dimensions, such social and mental well-being, in maintaining overall health. Neglecting these dimensions can have long-term negative consequences on an individual's well-being. In context traditional in-person therapy sessions, efforts are made to manually identify underlying factors that contribute disturbances, factors, if triggered, potentially lead severe health disorders. Our research focuses introducing a meticulous task aimed at identifying indicators wellness detecting their presence self-narrated human writings Reddit media platform. We mentioned Ethics Broader Impact.</p>

Language: Английский

Citations

1
Self-administered interventions based on natural language processing models for reducing depressive and anxious symptoms: Systematic review and meta-analysis (Preprint) DOI Creative Commons
David Villarreal‐Zegarra, Jackeline García-Serna, Gleni Quispe-Callo

et al.

Published: April 15, 2024

BACKGROUND The introduction of Natural Language Processing (NLP) technologies has significantly enhanced the potential self-directed interventions for treating anxiety and depression by improving human-computer interactions. Despite these advancements, particularly in AI Large Models (LLMs), robust evidence validating their effectiveness remains sparse. OBJECTIVE To determine whether based on NLP models can reduce depressive symptoms. METHODS Our study was a systematic review, protocol registered PROSPERO (CRD42023472120). databases we used review are Web Science, SCOPUS, MEDLINE (via PubMed), PsycINFO EBSCO), IEEE Xplore, EMBASE Cochrane Library. quality included studies assessed using JBI Critical Appraisal Tools. RESULTS 21 articles were selected 16 meta-analysis each outcome. overall showed that self-administered more effective reducing symptoms (SMD=0.819; 95%CI: 0.389-1.250; p<0.001) (SMD=0.272; 95% CI: 0.116-0.428; p=0.001) compared with various control conditions. In subgroup analysis, AI-based shown to be (SMD=1.059 [0.520 1.597]; (SMD=0.302 [0.073 0.532]; p=0.010) pooled Also, NLP-based outperform psychoeducation bibliotherapy both (SMD=1.481 [0.368 2.594]; p=0.009) (SMD=0.561 [0.195 0.927]; p=0.003). addition, than waitlist or no intervention anxious (SMD=0.196 [0.042 0.351]; p=0.013). CONCLUSIONS findings support usefulness self-applied alleviating widely prevalent mental health problems such as CLINICALTRIAL Protocol (CRD42023472120)

Language: Английский

Citations

0