Feasibility of Mental Health Triage Call Priority Prediction Using Machine Learning DOI Creative Commons
Rajib Rana, Niall Higgins, Kazi Nazmul Haque

et al.

Nursing Reports, Journal Year: 2024, Volume and Issue: 14(4), P. 4162 - 4172

Published: Dec. 20, 2024

Background: Optimum efficiency and responsiveness to callers of mental health helplines can only be achieved if call priority is accurately identified. Currently, operators making a triage assessment rely heavily on their clinical judgment experience. Due the significant morbidity mortality associated with illness, there an urgent need identify who have high level distress seen by clinician offer interventions for treatment. This study delves into potential using machine learning (ML) estimate from properties callers’ voices rather than evaluating spoken words. Method: Phone speech first isolated existing APIs, then features or representations are extracted raw speech. These fed series deep neural networks classify audio representation. Results: Development network architecture that instantly determines positive negative levels in input segments. A total 459 records helpline were investigated. The final ML model balanced accuracy 92% correct identification both instances priority. Conclusions: provides voice quality terms demeanor simultaneously displayed web interface computer smartphone.

Language: Английский

Artificial Intelligence and Decision-Making in Healthcare: A Thematic Analysis of a Systematic Review of Reviews DOI Creative Commons
Mohsen Khosravi, Zahra Zare,

Seyyed Morteza Mojtabaeian

et al.

Health Services Research and Managerial Epidemiology, Journal Year: 2024, Volume and Issue: 11

Published: Jan. 1, 2024

Introduction The use of artificial intelligence (AI), which can emulate human and enhance clinical results, has grown in healthcare decision-making due to the digitalization effects COVID-19 pandemic. purpose this study was determine scope applications AI tools process service delivery networks. Materials methods This used a qualitative method conduct systematic review existing reviews. Review articles published between 2000 2024 English-language were searched PubMed, Scopus, ProQuest, Cochrane databases. CASP (Critical Appraisal Skills Programme) Checklist for Systematic Reviews evaluate quality articles. Based on eligibility criteria, final selected data extraction done independently by 2 authors. Finally, thematic analysis approach analyze extracted from Results Of 14 219 identified records, 18 eligible included analysis, covered findings 669 other assessment score all reviewed high. And, 3 main themes including decision-making, organizational shared decision-making; originated 8 subthemes. Conclusions revealed that have been applied various aspects decision-making. improve quality, efficiency, effectiveness services providing accurate, timely, personalized information support Further research is needed explore best practices standards implementing

Language: Английский

Citations

47
Technology Readiness Assessment: Case of Clinical Decision Support Systems in Healthcare DOI

Oussama Laraichi,

Tugrul Daım, Saeed Alzahrani

et al.

Technology in Society, Journal Year: 2024, Volume and Issue: unknown, P. 102736 - 102736

Published: Oct. 1, 2024

Language: Английский

Citations

4
Les biais de l’IA : enjeux et précautions pour une prise de décision éthique et fiable en santé DOI
Édouard Lansiaux

Médecine de Catastrophe - Urgences Collectives, Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 1, 2025

Citations

0
Artificial Intelligence in Emergency Trauma Care: A Preliminary Scoping Review DOI Creative Commons
Christian Ventura,

Edward E Denton,

Jessica A. David

et al.

Medical Devices Evidence and Research, Journal Year: 2024, Volume and Issue: Volume 17, P. 191 - 211

Published: May 1, 2024

Abstract: This study aimed to analyze the use of generative artificial intelligence in emergency trauma care setting through a brief scoping review literature published between 2014 and 2024. An exploration NCBI repository was performed using search string selected keywords that returned N=87 results; articles met inclusion criteria (n=28) were reviewed analyzed. Heterogeneity sources explored identified by significance threshold P < 0.10 or an I 2 value exceeding 50%. If applicable, categorized within three primary domains: triage, diagnostics, treatment. Findings suggest CNNs demonstrate strong diagnostic performance for diverse traumatic injuries, but generalized integration requires expanded prospective multi-center validation. Injury scoring models currently experience calibration gaps mortality quantification lesion localization can undermine clinical utility permitting false negatives. Triage predictive now confront transparency, explainability, healthcare ecosystem barriers limiting real-world translation. The most significant gap centers on treatment-oriented AI applications provide real-time guidance urgent interventions rather than just analytical support. Keywords: intelligence, machine-learning, medicine, traumatology

Language: Английский

Citations

1
TriageIntelli: AI-Assisted Multimodal Triage System for Health Centers DOI Open Access

Ziad Araouchi,

Mehdi Adda

Procedia Computer Science, Journal Year: 2024, Volume and Issue: 251, P. 430 - 437

Published: Jan. 1, 2024

Language: Английский

Citations

1
Feasibility of Mental Health Triage Call Priority Prediction Using Machine Learning DOI Creative Commons
Rajib Rana, Niall Higgins, Kazi Nazmul Haque

et al.

Nursing Reports, Journal Year: 2024, Volume and Issue: 14(4), P. 4162 - 4172

Published: Dec. 20, 2024

Background: Optimum efficiency and responsiveness to callers of mental health helplines can only be achieved if call priority is accurately identified. Currently, operators making a triage assessment rely heavily on their clinical judgment experience. Due the significant morbidity mortality associated with illness, there an urgent need identify who have high level distress seen by clinician offer interventions for treatment. This study delves into potential using machine learning (ML) estimate from properties callers’ voices rather than evaluating spoken words. Method: Phone speech first isolated existing APIs, then features or representations are extracted raw speech. These fed series deep neural networks classify audio representation. Results: Development network architecture that instantly determines positive negative levels in input segments. A total 459 records helpline were investigated. The final ML model balanced accuracy 92% correct identification both instances priority. Conclusions: provides voice quality terms demeanor simultaneously displayed web interface computer smartphone.

Language: Английский

Citations

0