Purpose

The current study is a preliminary study to examine whether children with hearing loss would benefit from a speech recognition in noise training.

Methods

Twenty-five children who wore hearing aids, cochlear implants, or bimodal devices from 4 to 12 years old participated in the study (experimental, n = 16; control, n = 9). The experimental group received a speech-in-noise training that took sixteen 15-min sessions spanning 8 to 12 weeks. The task involves recognizing monosyllabic target words and sentence keywords with various contextual cues in a multi-talker babble. The target stimuli were spoken by two females and fixed at 65 dB SPL throughout the training while the masker varied adaptively. Pre- and post-training tests measured the speech recognition thresholds of monosyllabic words and sentences spoken by two males in the babble noise. The test targets were presented at 55, 65, and 80 dB SPL.

Results

The experimental group improved for word and sentence recognition in noise after training (Mean Difference = 2.4–2.5 dB, 2.7–4.2 dB, respectively). Training benefits were observed at trained (65 dB SPL) and untrained levels (55 and 80 dB SPL). The amount of post-training improvement was comparable between children using hearing aids and cochlear implants.

Conclusions

This preliminary study showed that children with hearing loss could benefit from a speech recognition in noise training that may fit into the children's school schedules. Training at a conversational level (65 dB SPL) transfers the benefit to levels 10–15 dB softer or louder. Training with female target talkers transfers the benefit to male target talkers. Overall, speech in noise training brings practical benefits for school-age children with hearing loss.

中文翻译：

---

听力损失儿童噪声训练中语音识别的初步研究

目的

目前的研究是一项初步研究，旨在检查听力损失儿童是否会从噪声训练中的语音识别中受益。

方法

25 名 4 至 12 岁佩戴助听器、人工耳蜗或双模装置的儿童参加了这项研究（实验组，n = 16；对照组，n = 9）。实验组接受了为期 8 至 12 周的 16 次 15 分钟的语音训练。该任务涉及在多说话者的胡言乱语中识别具有各种上下文线索的单音节目标词和句子关键字。目标刺激由两名女性说出，并在整个训练过程中固定在 65 dB SPL，而掩蔽器自适应变化。训练前和训练后测试测量了两名男性在混杂噪音中说出的单音节单词和句子的语音识别阈值。测试目标为 55、65 和 80 dB SPL。

结果

训练后，实验组在噪声中的单词和句子识别能力有所提高（平均差异分别为 2.4–2.5 dB、2.7–4.2 dB）。在受过训练的 (65 dB SPL) 和未受过训练的水平 (55 和 80 dB SPL) 下观察到训练益处。使用助听器和人工耳蜗的儿童在训练后的改善程度相当。

结论

这项初步研究表明，听力损失儿童可以从噪音训练中的语音识别中受益，这可能适合儿童的学校时间表。会话级别 (65 dB SPL) 的训练将益处转移到更柔和或更响亮的 10–15 dB 级别。与女性目标谈话者一起训练将好处转移给男性目标谈话者。总体而言，噪声训练中的语音为有听力损失的学龄儿童带来了实际好处。