EUVIMEDDescoperă EUVIMED
Mixed-Methods-StudieOpen access

Augmented Reality–Assisted Training Tool for Mental Health Task-Sharers: Pilot Mixed Methods Usability Study

Ling Li Vivian Ngiam iD, Lucas Piotr Wozniak iD, Catherine Dinh-Le iD, Ayanna Elon Seals iD, Victoria K Ngo iD, Jose Fernando Florez-Arango iD

JMIR XR and Spatial Computing · 2026-06-25 · 3 · e80711-e80711

DOI: 10.2196/80711 ↗

At a glance

Mixed-Methods-Studie in JMIR XR and Spatial Computing, published online on 2026-06-25. Written by Ling Li Vivian Ngiam and 5 co-authors. Institution: Neurohue, LLC, 600 N Broad Street, Suite 5 #3720, Middletown, DE, 19709, United States, 1 3013668555. 105 references. The full text is freely available in PubMed Central.

RezumatSubjectsDate bibliograficeCiteRelated studies

Abstract (original language)

Background

The growing global mental health (MH) burden, especially in underresourced communities, calls for innovative, scalable, and culturally responsive training approaches to expand care access and improve outcomes. Task-sharing has shown promise in addressing workforce shortages but is limited by training and supervision challenges. Traditional methods, such as role-playing and standardized patients, are resource-intensive and less scalable. Virtual simulations, including augmented reality (AR), present novel opportunities for immersive and interactive training. An AR-assisted training tool can enable culturally sensitive training while fostering empathy, communication skills, and confidence in handling nuanced MH scenarios. However, AR’s usability and effectiveness for MH task-sharing training remain underexplored.

Objective

This study aimed to assess the usability of an AR-assisted MH task-sharing training tool that uses virtual patient (VP) simulation and evaluate its potential to enhance training. Additionally, we developed design recommendations for future related XR-assisted clinical training tools.

Methods

We conducted a formative, explorative sequential mixed-methods usability study. A convenience sample of 5 MH trainees or workers (ages 18‐60 years; female: n=3, male: n=2; identifying as African American, Asian, and Hispanic) participated. Participants were recruited through a university-affiliated MH training program. The usability testing protocol included a semistructured prestudy interview, orientation to the AR headset (Magic Leap 2), a think-aloud user testing session, and a poststudy quantitative questionnaire and qualitative interview. Usability was assessed using a modified Post-Study System Usability Questionnaire (PSSUQ), which measures system usefulness, information quality, and interface quality. Data were analyzed using descriptive statistics and thematic analysis.

Results

The AR simulation was positively received by participants, demonstrating above-average usability. The overall mean PSSUQ score was 3.46 (SD 1.71), with subscale scores for system usefulness (mean 3.46, SD 1.77), information quality (mean 3.76, SD 1.73), and interface quality (mean 2.83, SD 1.59). Thematic analysis highlighted high realism fostered trainees’ empathy toward the VP, while increasing immersion and interaction quality. Despite some hardware limitations and user discomfort that broke immersion, participants recognized the tool’s potential and usefulness for training in various MH scenarios. Based on these findings, we proposed design recommendations across environmental context, training structure, VP behavioral realism (body language, voice, eye movement, and technical hardware considerations).

Conclusions

This pilot study is among the first to evaluate AR-based VP simulation for training lay MH task-sharers, filling the technology and population gaps of prior VR- and desktop-focused simulation research. Preliminary empirical usability evidence from a validated instrument (PSSUQ) demonstrates above-average usability, and findings informed design recommendations for future research. These findings suggest AR-based training could provide a less resource-intensive solution for realistic MH training practice in underresourced settings. Further research includes larger sample sizes for inferential analysis, comparison studies with traditional methods, and generalizing to other MH conditions.

Subjects

Author keywords

VP simulationaugmented realityaugmented reality glassesmental health care trainingparticipatory designrole-playing practicesimulation-based trainingsoft skills trainingtask-sharingusabilityvirtual standardized patient

Authors and institutions

  • Neurohue, LLC, 600 N Broad Street, Suite 5 #3720, Middletown, DE, 19709, United States, 1 3013668555.
  • Graduate School of Public Health and Health Policy, CUNY, New York, NY, United States.
  • Department Population Health Sciences, Weill Cornell Medicine, New York, NY, United States.
  • Center of Biomedical Informatics and Biostatistics (CB2), University of Arizona, 1230 North Cherry Avenue, PO Box 210242, Tuscon, AZ, 85721, United States, +1 (979) 481-7392.

ORCID profiles: Ling Li Vivian Ngiam, Lucas Piotr Wozniak, Catherine Dinh-Le, Ayanna Elon Seals, Victoria K Ngo, Jose Fernando Florez-Arango

Cite

APA

Ngiam, L. L. V., Wozniak, L. P., Dinh-Le, C., Seals, A. E., Ngo, V. K., & Florez-Arango, J. F. (2026). Augmented Reality–Assisted Training Tool for Mental Health Task-Sharers: Pilot Mixed Methods Usability Study. JMIR XR and Spatial Computing, 3, e80711-e80711. https://doi.org/10.2196/80711

Vancouver

Ngiam LLV, Wozniak LP, Dinh-Le C, Seals AE, Ngo VK, Florez-Arango JF. Augmented Reality–Assisted Training Tool for Mental Health Task-Sharers: Pilot Mixed Methods Usability Study. JMIR XR and Spatial Computing. 2026;3:e80711-e80711. doi:10.2196/80711

BibTeX
@article{Ngiam2026,
  author = {Ngiam, Ling Li Vivian and Wozniak, Lucas Piotr and Dinh-Le, Catherine and Seals, Ayanna Elon and Ngo, Victoria K and Florez-Arango, Jose Fernando},
  title = {Augmented Reality–Assisted Training Tool for Mental Health Task-Sharers: Pilot Mixed Methods Usability Study},
  journal = {JMIR XR and Spatial Computing},
  year = {2026},
  volume = {3},
  pages = {e80711--e80711},
  doi = {10.2196/80711}
}
Start for free →