听力学家引导的多模态人工智能用于纯音测听和鼓室压图解读与报告
Audiologist-Guided Multimodal AI for Pure-Tone Audiometry and Tympanometry Interpretation and Reporting.
文献信息
| PMID | 42821199 |
|---|---|
| 原文 | 在 PubMed 查看原文 ↗ |
| 发表日期 | 2026 |
| 作者 | Xia Wu |
| 作者单位 | Department of Otolaryngology-Head and Neck Surgery, Ningbo Hospital of Integrated Traditional Chinese and Western Medicine, Ningbo, China. |
| 期刊 | Journal of medical systems |
| SCI 分区 | Q1 |
| IF | 6 |
| 研究类型 | AI/ML · 临床 |
| 所属专科 | 耳科 |
中文摘要
多模态人工智能（AI）可能支持听力学报告，但端到端评分可能掩盖错误是源于图像转录、专科规则执行还是沟通。我们在155份去标识化门诊记录（代表151例患者）中将这些作为独立模块进行评估。在模块1中，在50份开发记录中比较了仅模式提示和听力学家技能提示；随后在101例独立患者中使用Codex GPT5.5智能体工作流测试锁定的技能。事后稳健性分析在相同的101例患者中，使用DeepSeek-V4-Flash-Vision-Exp测试了匹配的仅模式与技能辅助直接API条件。在开发阶段，该技能将听力损失类型一致性从85.0%提高到94.0%，将声反射一致性从70.4%提高到98.4%。Codex验证产生了1809/1820（99.4%）的精确数值阈值，但仅9/15（60.0%）的无反应条目正确。在DeepSeek稳健性分析中，该技能将类型一致性从37.6%提高到70.3%，而精确数值阈值一致性保持72.1%，无反应一致性保持0%。在模块2中，AI工作流接收的是经裁定的结构化值，而非模块1的输出。在两名听力学家使用60份记录完善报告后，评估了91例独立患者。所有25个规则衍生字段在91/91例Codex和87/91例DeepSeek病例中均正确；两名听力学家均将所有中文报告评为准确或基本准确。这些是不同的操作工作流，而非匹配的基础模型比较。这些发现支持对模块化、可审计的临床医生辅助进行前瞻性评估，而非自主诊断、端到端性能或患者可用性。
英文摘要
Multimodal artificial intelligence (AI) could support audiology reporting, but an end-to-end score can obscure whether errors arise from image transcription, specialty-rule execution, or communication. We evaluated these as separate modules in 155 de-identified outpatient records representing 151 patients. In Module 1, schema-only and audiologist-skill prompts were compared in 50 development records; the locked skill was then tested with a Codex GPT5.5 agentic workflow in 101 independent patients. A post hoc robustness analysis tested matched schema-only and skill-assisted direct-API conditions with DeepSeek-V4-Flash-Vision-Exp in the same 101 patients. The skill increased hearing-loss-type agreement from 85.0% to 94.0% and acoustic-reflex agreement from 70.4% to 98.4% in development. Codex validation yielded 1809/1820 (99.4%) exact numeric thresholds, but only 9/15 (60.0%) no-response entries were correct. In the DeepSeek robustness analysis, the skill increased type agreement from 37.6% to 70.3%, while exact numeric-threshold agreement remained 72.1% and no-response agreement remained 0%. In Module 2, AI workflows received adjudicated structured values, not Module 1 outputs. After two audiologists refined reports using 60 records, 91 independent patients were evaluated. All 25 rule-derived fields were correct in 91/91 Codex and 87/91 DeepSeek cases; both audiologists rated all Chinese reports accurate or basically accurate. These were different operational workflows, not a matched foundation-model comparison. The findings support prospective evaluation of modular, auditable clinician assistance, not autonomous diagnosis, end-to-end performance, or patient usability.