摘要
针对传统耳鸣声治疗方法在患者掩蔽频段的对应效果不精确或掩蔽效果不佳的问题,本研究基于时频图表示(time-frequency representations,TFR)方法,提出了一种多轨个性化耳鸣康复音(multi-track personalized tinnitus rehabilitation sound,MT-PTRS)的生成方法。该方法不仅能体现耳鸣掩蔽疗法所需的频率集中特性,且一定程度上提高了耳鸣康复音的可听性。本研究基于基本的自然声库,通过TFR方法识别出自然声中的主要能量频段,并结合患者耳鸣频段对自然声进行切割、分类,进一步合成频率集中的个性化自然掩蔽声。本研究将个性化自然掩蔽声作为主要音轨,结合宽带自然声生成MT-PTRS,相较传统掩蔽治疗中的窄带噪声,具有更好的可听性和自然性。实验结果表明,相较于其他的多轨耳鸣康复音方法,该方法生成的康复音的悦耳度和丰富度有一定提升,增加了患者对声治疗的耐受性,对耳鸣声治疗研究具有一定参考价值。
To address the issues of inaccurate correspondence or poor masking effects of traditional tinnitus sound therapy methods in the masked frequency range of patients.A synthesis method for generating multi-track personalized tinnitus rehabilitation sound(MT-PTRS)was proposed based on the time-frequency representation(TFR)method,which not only reflected the characteristics of frequency concentration,but also improved the audibility of tinnitus rehabilitation sounds to some extent.Based on a basic natural sound library,the TFR method was used to identify the main energy frequency bands in the natural sounds.Combined with the tinnitus frequency range of the patients,the natural sounds were segmented and classified to further synthesize personalized natural masking sounds with frequency concentration.This study combined personalized natural masking sounds as the main track with broadband natural sound to generate MT-PTRS.MT-PTRS had better audibility and naturalness than narrow-band noise in traditional masking therapy.Compared with the previous multi-track tinnitus rehabilitation sound methods,the rehabilitation sound generated by the proposed method had certain improvement in pleasantness and richness,which increased the patient′s tolerance to sound therapy.This study has certain reference value for the research of tinnitus sound treatment.
作者
谢子奇
何培宇
陈娅南
李永康
方安成
潘帆
XIE Ziqi;HE Peiyu;CHEN Yanan;LI Yongkang;FANG Ancheng;PAN Fan(College of Electronics and Information Engineering,Sichuan University,Chengdu 610065,China)
出处
《生物医学工程研究》
2024年第2期115-122,共8页
Journal Of Biomedical Engineering Research
基金
四川省自然科学基金项目(2022NSFSC0799)。
关键词
耳鸣声治疗
时频图表示
频率匹配
声音合成
自然音
Tinnitus sound therapy
Time-frequency representations
Frequency matching
Sound synthesis
Natural sound
