甲状腺切除术后早期嗓音变化的多模态模式

文献信息

中文摘要

目的： 甲状腺切除术后即使没有明显的声带麻痹，也可能出现嗓音变化。本研究探讨了术后早期在患者报告、声学、功能和频闪喉镜测量方面的变化模式。
方法： 这项单中心回顾性研究纳入120例接受甲状腺切除术并在术后第7±2天完成术前和术后早期嗓音评估的患者。排除既往有嗓音疾病或明显声带麻痹的患者。评估包括甲状腺切除术相关嗓音和症状问卷、声学分析、最长发声时间、嗓音范围图和电视喉频闪镜。术后变化变量经方向性统一、标准化后，归纳为症状、声学、功能和频闪喉镜领域评分，用于探索性K均值聚类。
结果： 探索性三聚类方案包括：A聚类，相对恶化较轻（66例，55.0%）；B聚类，声学和频闪喉镜明显恶化（8例，6.7%）；C聚类，症状和功能恶化而声学变化有限（46例，38.3%）。A聚类相对队列平均值恶化较轻，但症状和音域也出现恶化。B聚类在微扰（shimmer）、噪声谐波比、水肿和声带弓形方面恶化最明显。C聚类症状和音域恶化突出，而噪声谐波比和声带弓形变化相对较小。女性与B/C合并归属的探索性关联并不能确立生物学易感性。
结论： 探索性聚类描述了甲状腺切除术后早期嗓音变化在常规可获得的临床嗓音测量中的不同特征。这些特征应被解释为描述性的术后早期模式，而非固定的诊断或预后分类。尚需进一步的纵向研究和外部验证。

英文摘要

OBJECTIVES: Voice change after thyroidectomy may occur even without overt vocal fold paralysis. This study explored patterns of early postoperative change across patient-reported, acoustic, functional, and stroboscopic measures.
METHODS: This single-center retrospective study included 120 patients who underwent thyroidectomy and completed preoperative and early postoperative voice assessments at postoperative day 7 ± 2 days. Patients with preexisting voice disorders or overt vocal fold paralysis were excluded. Assessments included the thyroidectomy-related voice and symptom questionnaire, acoustic analysis, maximum phonation time, voice range profile, and videolaryngostroboscopy. Postoperative change variables were directionally harmonized, standardized, and summarized into symptom, acoustic, functional, and stroboscopic domain scores for exploratory K-means clustering.
RESULTS: An exploratory three-cluster solution comprised cluster A, with lower relative deterioration (66 patients, 55.0%); cluster B, with marked acoustic and stroboscopic deterioration (8 patients, 6.7%); and cluster C, with symptom and functional deterioration with limited acoustic change (46 patients, 38.3%). Cluster A showed less deterioration relative to the cohort average, although symptoms and pitch range also worsened. Cluster B showed the greatest worsening in shimmer, noise-to-harmonic ratio, edema, and bowing. Cluster C showed prominent symptom and pitch-range deterioration with comparatively small changes in noise-to-harmonic ratio and bowing. The exploratory association of female sex with combined B/C membership does not establish biological susceptibility.
CONCLUSION: Exploratory clustering described differing profiles of early post-thyroidectomy voice change across routinely obtainable clinical voice measures. These profiles should be interpreted as descriptive early postoperative patterns rather than fixed diagnostic or prognostic categories. Further longitudinal and external validation is needed.