机构地区:[1]解放军总医院耳鼻咽喉头颈外科医学部、国家耳鼻咽喉疾病临床医学研究中心,北京100853 [2]浙江医院耳鼻咽喉科,杭州310007 [3]浙江中医药大学医学技术学院,杭州310053 [4]首都医科大学生物医学工程学院,北京100069 [5]北京语言大学语言康复学院,北京100083 [6]俄亥俄大学,俄亥俄州阿森斯45701,美国
出 处:《中华耳鼻咽喉头颈外科杂志》2023年第11期1112-1122,共11页Chinese Journal of Otorhinolaryngology Head and Neck Surgery
基 金:国家重点研发计划(2020YFC2004005);国家自然科学基金(61370023,81460099)。
摘 要:目的为克服现有中文言语测听材料在成年耳蜗植入者中的天花板效应,依照英文AzBio句表的开发流程,开发一套更难的、更贴近现实生活的中文普通话语句测听表,并在成人耳蜗植入者验证其等价性。方法选取成人日常生活口语素材,邀请2男2女播音专业学生在本底噪声低于20 dB(A)的隔声室内录音1020句,以声码器技术仿真成5通道人工耳蜗的音效,编撰成51张表,以拉丁方设计在17名听力正常人中进行言语识别率测试,从均值±1标准差(x±1s,61.9%至92.2%)的范围内挑选出同质性良好的600句(每个播音人150句,识别率均值78.0%),编撰成每表20句共30张CMnBio句表。另行招募30例使用年限12个月~12年不等的成年人工耳蜗植入者,在65 dBSPL声级下依序测试30张表的言语识别率。采用拉丁方设计方案,每张测试表都有同样的机会成为首张测试表,以避免练习效应和疲劳效应的干扰。使用SPSS 18.0统计软件进行数据分析。结果(1)30例人工耳蜗植入者的言语识别率呈现出较大的个体间差异,但每例植入者对30张表的言语识别率相对稳定。(2)对言语识别率进行合理化的反正弦变换后,计算每例植入者每表识别率与其30张表平均识别率的差值;以30张CMnBio表作为考察对象,对30例植入者的上述差值进行重复测量的方差分析,F=3.5031,P<0.001;进而采用Post Hoc Tukey两两比较,除第4、5、10、29号表外,其余26张表之间差异无统计学意义(P>0.05)。(3)26张等价CMnBio句表在30例成年人工耳蜗植入者中的识别率为(65.4±13.3)%,呈良好的正态分布。(4)采用n=65的二项分布模型可最好地反映表间的变异性。结论本研究建立了面向成年人工耳蜗植入者、具有良好等价性的26张普通话CMnBio句表,并提供了使用单张表和二张表时识别率得分95%置信度下的临界差值。为普通话成人人工耳蜗临床实践及开展跨语种成效对比研究,提供了一个Objectives The goal of the present study was to develop and validate a set of Mandarin Chinese adaptation of AzBio sentence test(CMnBio)used for the assessment of speech recognition in hearing-impaired listeners and cochlear implant(CI)users.Methords Following the procedures of development of the English AzBio sentence materials,the present study was conducted in two stages.In the first stage,a total of 1850 sentences were compiled and recorded by four Mandarin-speaking adult talkers(two males and two females),from which 1020 sentences with better sound quality and higher naturalness were retained.All the 1020 sentences were processed through a 5-channel noise vocoder and presented to 17 normal-hearing Mandarin-speaking adults for recognition.A total of 600 sentences(150 from each talker)in the range of approximately 62% to 92% correct were subsequently selected to compile 30,20-sentence lists.In the second stage,30 postlingually-deafened adult CI users were recruited to verify the list equivalency.Finally,the binomial distribution model was adopted to account for the inherent variability in the lists and to generate the lower and upper limits of the 95% critical differences.Results The average sentence recognition of the 5-channel vocoder-processed 30 lists of Mandarin Chinese AzBio sentences(600 sentences total)by the 17 normal-hearing listeners was 78.0% correct.The validation based on the performance scores of the 30 CI participants revealed that 26 of the 30 lists were equivalent through a repeated-measures analysis of variance and a following post hoc Tukey′s test.The binomial distribution modelling indicated that the inter-list variability could be accounted for with a 65-item binomial distribution model.The lower and upper limits of the 95% critical differences for one-and two-list recognition scores were generated.Conclusions The Mandarin Chinese adaptation of AzBio sentence test(CMnBio)is developed and validated for speech recognition assessment of CI users.The final set of 26,20-sentence lists shows
关 键 词:测听法 言语 语句识别 耳蜗植入术 语后聋 汉语普通话
分 类 号:R764.9[医药卫生—耳鼻咽喉科]
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...