耳蜗机制与耳声发射测试性能
研究纳入212名听力正常或轻中度听力损失受试者,比较控制耳蜗来源机制对畸变产物耳声发射(DPOAE)识别听力状态准确性的影响。结果显示,在整体人群中平滑DPOAE精细结构以减少反射源贡献并未提高测试准确性;仅在听力正常且精细结构确认的亚组中,低刺激强度下准确性有小幅提升(2%-5%)。
要点
- 整体人群中,减少反射源贡献未改善DPOAE测试准确性。
- 听力正常且精细结构确认者中,低强度下准确性提升2%-5%。
- 反射源AM-SFOAE未比任何DPOAE条件更准确识别听力损失。
核心结论
- 在整体人群中,通过平滑DPOAE精细结构限制耳蜗来源贡献不会提高DPOAE测试准确性。
- 使用AM-SFOAE技术记录反射源耳声发射不会比混合源或单源DPOAE更准确识别听力状态。
研究设计与样本
- 研究纳入212名听力正常或轻中度听力损失受试者,根据DPOAE筛查方案分为不确定识别组和确定识别组。
- 记录f2=1、2、4 kHz周围的DPOAE精细结构(f2/f1比=1.22),L2=35、45、55 dB SPL(L1/L2比=10 dB)。
- 使用离散余弦变换平滑精细结构,并记录调幅刺激频率耳声发射(AM-SFOAE)。
- 以相对操作特征曲线下面积(AROC)量化测试准确性,并评估听力正常耳固定5%失败率下的表现。
主要发现
- 在整体受试者中,减少反射源贡献至DPOAE未在任何测试频率产生优于最佳控制条件的测试性能。
- 在听力正常且精细结构确认的亚组中,减少反射源贡献使识别准确性相对最佳控制条件出现2%-5%的小幅提升,见于最低刺激强度(L2=35或45 dB SPL)。
- 在该亚组中,低L2下畸变源DPOAE在听力正常耳5%固定假阳性率下更准确识别轻度听力损失,而这些条件在较大群体中准确性较差。
- 控制来源贡献对中度听力损失识别的影响在缩减亚组中不明确,部分条件畸变源DPOAE更准确,部分条件无变化。
- 无论在大群体还是缩减群体中,反射源AM-SFOAE均未比任何DPOAE条件更准确识别听力损失耳。
关键数值与推荐
- DPOAE精细结构记录参数:f2/f1比=1.22,L2=35、45、55 dB SPL,L1/L2比=10 dB。
- 听力正常耳失败率固定为5%用于评估不同行为阈值类别的识别准确性。
- 准确性量化指标为相对操作特征曲线下面积(AROC)。
临床怎么用
- 在一般人群的DPOAE测试中,无需通过平滑精细结构来限制耳蜗来源贡献以期望提高准确性。
- 对于听力正常且精细结构确认的个体,低刺激强度下畸变源DPOAE可能小幅提升轻度听力损失的识别准确性。
- AM-SFOAE技术记录反射源耳声发射并不优于DPOAE,不建议替代DPOAE用于听力状态识别。
局限性
- 研究仅纳入听力正常或轻中度听力损失受试者,未涵盖重度或极重度听力损失。
- 结果在缩减亚组中部分条件样本量可能有限,影响统计效能。
- 未报告受试者年龄分布、性别比例等人口学特征,可能影响外推性。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
OBJECTIVES: This study aims to determine the impact of controlling cochlear-source mechanism on the accuracy with which auditory status is identified using otoacoustic emissions (OAEs) in two groups of subjects with normal hearing (NH) and subjects with mild to moderate hearing loss. DESIGN: Data were collected from 212 subjects with NH and with mild to moderate hearing loss who fell into two categories based on a distortion product OAE (DPOAE) screening protocol: the uncertain-identification group (where errors were likely) and the certain-identification group (where errors were unlikely). DPOAE fine-structure patterns were recorded at intervals surrounding f2 = 1, 2 and 4 kHz (f2/f1 ratio = 1.22), with L2 = 35, 45, and 55 dB SPL (L1/L2 ratio = 10 dB). The discrete cosine transform was used to smooth fine structure, limiting the source contribution to the distortion source only. Reflection-source OAEs were also recorded using amplitude-modulated stimulus frequency OAEs (AM-SFOAE). Area under the relative operating characteristic (AROC) curve was used to quantify test accuracy when the source contribution was controlled versus the condition where both sources contribute. Additionally, failure rate, fixed at 5% for NH ears, as a function of behavioral-threshold category was evaluated. RESULTS: When data for the entire subject group were examined, reducing the reflection-source contribution to the DPOAE did not result in better test performance than the best control condition at any frequency tested. When the subjects with NH were restricted to those with confirmed fine structure, AROC analyses indicated that reducing the reflection-source contribution resulted in several small increases in the accuracy (2%-5%) with which auditory status was identified relative to the best control condition. This improvement was observed for the lowest stimulus levels (i.e., L2 = 35 or 45 dB SPL). In this subset of subjects, distortion-source DPOAEs resulted in more accurate identification of mild hearing loss for a fixed false-positive rate of 5% in NH ears at lower L2's, conditions with poor accuracy in the larger group of subjects. The impact of controlling the source contribution on the identification of moderate losses was less clear in the reduced subject group, with some conditions where the distortion-source DPOAE was more accurate than the control condition and other conditions where there was no change. There was no evidence that reflection-source AM-SFOAEs more accurately identified ears with hearing loss when compared to any of the DPOAE conditions in either the large or reduced group of subjects. CONCLUSION: While improvements in test accuracy were observed for some subjects and some conditions (e.g., mild hearing losses and low stimulus levels in the reduced subset of subjects), these results suggest that restricting cochlear source contribution by "smoothing" DPOAE fine structure is not expected to improve DPOAE test accuracy in a general population of subjects. Likewise, recording reflection-source OAEs using the AM-SFOAE technique would not be expected to more accurately identify hearing status compared to mixed- or single-source DPOAEs.
引用
- Go NA, Stamper GC, Johnson TA, et al. Cochlear Mechanisms and Otoacoustic Emission Test Performance. Ear Hear. 2019. PMID: 29952805. doi:10.1097/AUD.0000000000000625
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。