METHODS: The 1226 audiograms of hearing-impaired individuals aged 60-90 years were reviewed. The random sample function randomly assigned 613 participants to the training and testing sets each. A linear model was created to predict the PTT value at each frequency based on variables significant at all frequencies across 0.5-4 kHz. The adjusted-R2 value was considered to indicate the performance of the predictive model. Pearson\'s correlation coefficient was used to describe the relationship between the actual and predicted PTT at 0.5, 1, 2, and 4 kHz among the testing set to measure the performance of the proposed model.
RESULTS: The predictive model was devised using variables based on the speech recognition threshold (SRT) after adjusting with age in the training set. The overall prediction accuracy demonstrated a higher adjusted-R2 ranging from 0.74 to 0.89 at frequencies of 0.5, 1, and 2 kHz, whereas a low percentage of explained variance was observed at 4 kHz (adjusted-R2 = 0.41). This predictive model can serve as an adjunctive clinical tool for guiding determination of the PTTs. Moreover, the predicted PTTs can be applied in the hearing aid programming software to set appropriate hearing aid gain using standard prescriptive formulas.
方法:回顾了年龄在60-90岁的听力受损者的1226个听力图。随机样本函数将613名参与者随机分配到训练集和测试集。创建线性模型以基于在跨0.5-4kHz的所有频率处显著的变量来预测在每个频率处的PTT值。调整后的R2值被认为表明了预测模型的性能。Pearson的相关系数用于描述测试集之间在0.5、1、2和4kHz的实际和预测的PTT之间的关系,以衡量所提出模型的性能。
结果:在训练集中根据年龄进行调整后,使用基于语音识别阈值(SRT)的变量设计了预测模型。在0.5、1和2kHz的频率下,总体预测精度显示出更高的调整R2,范围从0.74到0.89。而在4kHz时观察到低百分比的解释方差(调整后的R2=0.41).该预测模型可以用作指导确定PTT的辅助临床工具。此外,可以在助听器编程软件中应用预测的PTT,以使用标准处方公式来设置适当的助听器增益。