AI-based communication training system with speech recognition, SEGUE-aligned semantic analysis, and automated feedback reports¶

Design · All designs
Evidence · 2 claims (2 for) · 1 study (1 causal), q3 · 1 of 1 report an effect size · 2 claims rest on one study

Description¶

A web-based software system built on the DeepSeek-chat large language model (R1 version, accessed via cloud API, not fine-tuned) that simulates patients for SEGUE-based communication training. It "integrated an automatic speech recognition engine (e.g., iFlytek) with a semantic analysis module aligned with the five SEGUE dimensions", transcribes spoken dialogue, scores SEGUE behaviors, and generates a structured post-encounter feedback report with dimension scores and example response formulations. Interaction is text/voice only, without visual avatars, and the AI-VSP acts solely as a patient without instructional prompts.

Design Implications¶

Context¶

Requirements¶

Stable internet connectivity for cloud-based model access
Structured prompt engineering with rule-based SEGUE constraints and manual iterative adjustment validated by expert voting

Constraints¶

No visual avatars, facial animation, or synthesized imagery; lacks nonverbal and affective variability of human SPs
Case materials restricted to a limited set of cases (10 per SP)

Target Learners¶

Senior undergraduate clinical medicine students

Learning Goals¶

Physician-patient communication behaviors assessed by the SEGUE framework

Claims¶

Ai Vsp More Training Cycles [+M]
Ai Vsp Higher Total Content Segue Scores [+M]

Examples¶

-

Key Sources¶

Sun Ning, Zhou Xintong, Yang Zhongqian, Zhou Yu, Ma Ruihang, Wang Zhong. (2026). Application of AI-based virtual standardized patients in physician-patient communication training: a study based on the SEGUE framework. Frontiers in Public Health. https://doi.org/10.3389/fpubh.2026.1768518

Page metadata

Field	Value
`type`	`design`
`id`	`ai-vsp-segue-training-system`
`title`	`AI-based communication training system with speech recognition, SEGUE-aligned semantic analysis, and automated feedback reports`
`description`	`A web-based software system built on the DeepSeek-chat large language model (R1 version, accessed via cloud API, not fine-tuned) that simulates patients for SEGUE-based communication training.`
`status`	`review`
`generated`	by `process:settle-candidates` on `2026-11-05`
`trust`	unverified — no one has confirmed this page's content
`sources`	1 — `sun-ning-2026`

Frontmatter is the page's machine-readable record — see the Schema & Guide.