听觉大语言模型辅助医生在喉镜检查前判断喉部疾病
Clinician Assessment of Laryngeal Disease with Auditory AI before Laryngoscopy: A Paired Reader Study
医生只听声音就能在喉镜前初筛喉部病变,这篇研究用 32 名医生、800 次配对评估量化了收益和风险:癌前病变检出敏感度从 53.9% 涨到 75.0%,但模型出错时医生也会被带偏。做医疗 AI 的开发者值得看看这套评估卡设计和读者研究方法。
一项发表于 arXiv 的配对阅片研究测试了听觉大语言模型在喉镜检查前的作用。模型基于语音录音、患者信息和声学指标生成评估卡,含初步分类、嗓音质量评分、理由和恶性病变预警。在 800 次配对评估中,32 名耳鼻喉科医生的分类准确率从 49.8% 提升至 56.8%,提高 7.0 个百分点(P < 0.001);模型判断正确时医生准确率上升 14.7 个百分点,错误时下降 8.7 个百分点。
模型本身的初步分类准确率为 72%(保留测试集)和 67%(时间验证队列),恶性预警对癌前或恶性病变的检出率为 85.7% 和 86.8%。借助 AI 复核后,医生对癌前或恶性病变的敏感度从 53.9% 提高到 75.0%,提升 21.1 个百分点。
研究局限在于单中心数据和序列设计无法完全区分 AI 辅助与二次评估效应,且医生未能有选择地采纳模型判断。作者指出需要在随机化设计和外部数据集上进一步验证恶性预警功能。
Clinician Assessment of Laryngeal Disease with Auditory AI before Laryngoscopy: A Paired Reader Study
Background: Before laryngoscopy, clinicians assess the likely cause of hoarseness and whether to suspect a precancerous or malignant lesion. Whether voice-based artificial intelligence (AI) supports pre-laryngoscopic categorization is unknown. Methods: We developed an auditory large language model that generates a patient-level assessment card (provisional category, voice-quality ratings, and rationale) and a prespecified malignancy alert from voice recordings, patient information, and acoustic measures obtained before laryngoscopy at one center. We evaluated both in a held-out test set and a later cohort. In a sequential reader study, 32 otolaryngology clinicians categorized held-out cases without AI and, 1 to 2 weeks later, reviewed their initial categories with the card (without the alert). Results: Across 800 paired assessments, reader accuracy was 49.8% before and 56.8% after review (difference, 7.0 percentage points; 95% CI, 1.7 to 13.5; P < 0.001); 99 categories changed from incorrect to correct and 43 from correct to incorrect (ratio, 2.30; 95% CI, 1.22 to 5.31). Accuracy rose 14.7 points when the model was correct and fell 8.7 when incorrect. Provisional-category accuracy was 72% in held-out testing and 67% in temporal validation. The malignancy alert flagged 85.7% and 86.8% of precancerous or malignant lesions. Sensitivity for precancerous or malignant lesions increased from 53.9% to 75.0% after AI-assisted review (difference, 21.1 percentage points; 95% CI, 4.3 to 42.0). Conclusions: In this single-center study, clinicians' pre-laryngoscopic categorization accuracy was higher after reviewing the AI assessment card, although the sequential design cannot fully separate AI assistance from reassessment; gains depended on model correctness without clearly selective adoption. Randomized and external evaluations, including of the malignancy alert, are needed.
本条情报归属 AITOP「论文」核心栏目,重点聚焦北美前沿实验室与硅谷科技巨头。在跨区域产业观察坐标中,该动态直接关联「AI 开发者工具链、Coding Agent 与工程化实践」研究主线,反映出当前技术演进与商业落地的最新风向。涉及核心实体与技术议题包括:听觉大语言模型、医疗AI、喉部疾病、读者研究,读者可通过下方标签进入对应专题页查阅历史脉络与关联对比。
信源核验说明:本文原始线索采编自「arXiv cs.AI」(@Siyuan Hou, Shuting Yang, Zengrui Jin, Liyu Cheng, Siyin Wang, Yunhan Shi, Xueyan Li, Rong Hu, Yuhong Lin, Qiyu Chen, Yanru Li, Ji Wu, Bowen Zhou, Mangsuo Zhao, Wen Xu, Chao Zhang),经 AITOP 多维评分管线结构化提炼。