基于递归图分析的腭裂语音鼻漏气自动识别算法

doi:10.11896/jsjkx.181001848

计算机科学 ›› 2020, Vol. 47 ›› Issue (2): 95-101.doi: 10.11896/jsjkx.181001848

所属专题：医学图像

• 计算机图形学&多媒体 • 上一篇下一篇

基于递归图分析的腭裂语音鼻漏气自动识别算法

刘新怡¹,田维维¹,梁文茹¹,何凌¹,尹恒²

(四川大学电气信息学院成都610065)¹;
(口腔疾病研究国家重点实验室成都610041)²

收稿日期:2018-10-06 出版日期:2020-02-15 发布日期:2020-03-18
通讯作者: 尹恒(yinheng@scu.edu.cn)
基金资助:
国家自然科学基金青年科学基金(61503264)

Automatic Detection Algorithm of Nasal Leak in Cleft Palate Speech Based on Recursive Plot Analysis

LIU Xin-yi¹,TIAN Wei-wei¹,LIANG Wen-ru¹,HE Ling¹,YIN Heng²

(College of Electrical Engineering and Information Technology,Sichuan University,Chengdu 610065,China)¹;
(State Key Laboratory of Oral Diseases,Chengdu 610041,China)²

Received:2018-10-06 Online:2020-02-15 Published:2020-03-18
About author:LIU Xin-yi,born in 1997,postgraduate.Her main research interests include Speech signal processing;YIN Heng,born in 1971,master.Her main research interests include cleft palate speech assessment.
Supported by:
This work was supported by the Young Scientists Fund of the National Natural Science Foundation of China (61503264).

摘要/Abstract

摘要： 鼻漏气是腭咽闭合不全患者的典型症状,针对腭裂语音鼻漏气的特征进行研究,利用基于非线性动力学方法的递归图对特征进行发掘,并结合递归趋势分析法和基于递归图的区域进行分布处理,提取递归图分析的量化参数和最小区域矩阵作为特征参数。结合分类器,实现对腭裂语音鼻漏气的自动识别。实验针对降采样点、延迟时间、临界距离、语音单元、分类器种类等因素,进行了识别效果的分析,并综合权衡各因素对识别正确率的影响,选取了最优取值。实验结果表明,采用KNN分类器并当降采样点为30000点、延迟时间为3ms、临界距离5个单位、语音单元为4帧时,腭裂语音鼻漏气自动识别的正确率达84.63%。腭裂语音鼻漏气自动识别算法能为临床腭咽功能评估提供高效、客观的辅助诊断依据。

关键词: 鼻漏气, 递归图, 腭裂语音, 区域分布

Abstract: Nasal leak is a typical symptom of patients with velopharyngeal insufficiency.This paper studied the characteristics of nasal leak in cleft palate speech.Recursive plot based on the nonlinear dynamics method is used to explore the features.Combined with the recursive trend analysis method and the region distribution processing based on the recursive plot, quantitative parameters and minimum regions of the recursive plot analysis are extracted as characteristic matrix.Combined with classifier,automatic detection of nasal leak in cleft palate speech is achieved.The experiment analyzes the detection effect for factors such as downsampling point,delay time,critical distance,speech unit and classifier type then comprehensively weighs the influence of each factor on the detection accuracy in order to select the optimal value.The experimental results show that when the KNN classifier is used,the downsampling point is 30000 points,the delay time is 3ms,the critical distance is 5 units,and the speech unit is 4 frames,the detection accuracy of nasal leak in cleft palate speech is 84.63%.The automatic detection algorithm of nasal leak in cleft palate speech is aimed at providing an effective and objective auxiliary diagnosis basis for clinical pharyngeal function assessment.

Key words: Cleft palate speech, Nasal leak, Recursive plot, Region distribution

中图分类号:

TP391.9

刘新怡,田维维,梁文茹,何凌,尹恒. 基于递归图分析的腭裂语音鼻漏气自动识别算法[J]. 计算机科学, 2020, 47(2): 95-101. https://doi.org/10.11896/jsjkx.181001848

LIU Xin-yi,TIAN Wei-wei,LIANG Wen-ru,HE Ling,YIN Heng. Automatic Detection Algorithm of Nasal Leak in Cleft Palate Speech Based on Recursive Plot Analysis[J]. Computer Science, 2020, 47(2): 95-101. https://doi.org/10.11896/jsjkx.181001848

参考文献

[1]雷丽.腭裂语音治疗学(修订1版)[M].武汉:湖北科学技术出版社,2004.
[2]HE L,ZHANG J,LIU Q,et al.Automatic detection of glottal stop in cleft palate speech[J].Biomedical Signal Processing and Control,2018,39:230-236.
[3]GLADE R S,DEAL R.Diagnosis and Management of Velopharyngeal Dysfunction[J].Oral & Maxillofacial Surgery Clinics of North America,2016,28(2):181-188.
[4]DHAKY K,BULSARA M,SETHNA B.Speech Therapy and Assessment (via Multimedia Devices for Cleft Lip and Palate Patients)[C]∥IEEE Global Humanitarian Technology Conference.Washington: IEEE Computer Society,2011:415-418.
[5]SAMOY K,HENS G,VERDONCK A,et al.Surgery for velopharyngeal insufficiency:The outcomes of the University Hospitals Leuven [J].International Journal of Pediatric Otorhinola-ryngology,2015,79(12):2213-2220.
[6]WANG G M,YUAN W H,WARREN D W.Evaluation of the function of pronunciation aid by air pressure and airflow measu-ring instrument[J].Shanghai Journal of Stomatology,1998,7(2):13-16.
[7]SHEN T B,MENG X Y.Comparative study of velopharyngeal closure before and after operation in children with cleft palate[J].Journal of Modern Integrated Traditional Chinese and Western Medicine,2010,19(3):319-320.
[8]LIU G,PU G C,LI W S.Effect of early cleft palate repair on velopharyngeal closure in patients[J].Chongqing Medical Journal,2015,44(34):4805-4806,4809.
[9]WANG G M.Diagnosis and treatment of pharyngeal closure dysfunction [J].Journal of Oral and Maxillofacial Surgery,2003,13(4):339-345.
[10]LEI L.Evaluation of pharyngeal dysfunction by nasopharyngeal fiberoscopy[C]∥Chinese Journal of Stomatology.FDI World Stomatological Congress Abstracts.Chinese Journal of Stomatology,2006:2.
[11]RAH D K,KO Y L,LEE C,et al.A noninvasive estimation of hypernasality using a linear predictive model[J].Annals of Biomedical Engineering,2001,29(7):587-594.
[12]AKAFI E,VALI M,MORADI N.Detection of hypernasal speech in children with cleft palate[C]∥Biomedical Engineering.IEEE,2013:237-241.
[13]CASTELLANOS G,DAZA G,SÁNCHEZ L,et al.Acoustic speech analysis for hypernasality detection in children[C]∥International Conference of the IEEE Engineering in Medicine & Biology Society.IEEE,2006:5507-5510.
[14]DUBEY A K,PRASANNA S R M,DANDAPAT S.Zero time windowing analysis of hypernasality in speech of Cleft Lip and palate children[C]∥Communication.IEEE,2016:1-6.
[15]OROZCO-ARROYAVE J,BELALCAZAR-BOLANOS E, ARIAS-LONDONO J,et al.Characterization methods for the detection of multiple voicedisorders:neurological,functional,and organic diseases[J].IEEE Journal of Biomedical & Health Informatics,2015,19(6):1820-1828.
[16]NIETO R G,MARNÍ-HURTADO J I,CAPACHO-VALBUENA L M,et al.Pattern recognition of hypernasality in voice of patients with Cleft and Lip Palate[C]∥2014 XIX Symposium on Image,Signal Processing and Artificial Vision.IEEE,2014:1-5.
[17]CRUZ C D L,SANTHANAM B.A joint EMD and teager-kaiser energy approach towards normal and nasal speech analysis[C]∥Conference on Signals,Systems and Computers.IEEE,2017.
[18]BELALCAZAR-BOLAÑOS E,VILLA-CAÑAS T,BEDOYA-JARAMILLO S,et al.Feature selection for hypernasality detection using PCA,LDA,kernel PCA and greedy kernel PCA[C]∥Image,Signal Processing,and Artificial Vision.IEEE,2012:246-251.
[19]YAN R Q.Recursive analysis of speech signal dynamics characteristics [D].Shanghai:Shanghai Jiaotong University,2006.
[20]YAN R Q,ZHU Y S.Unvoiced and voiced decision based on quantitative recursive analysis[J].Journal of Electronics & Information Technology,2007(7):1703-1706.
[21]JIA L,YIN Y,YANG H C.Application of recursive analysis in the detection of noisy speech endpoints[J].Journal of Shenyang Aerospace University,2017,34(6):83-86.
[22]YAN R Q,ZHU Y S.Speech endpoint detection method based on signal recursion analysis[J].Journal on Communications,2007,28(1):35-39.
[23]LI J,WANG J F,GAO J D.Speech endpoint detection algorithm based on empirical mode decomposition and recursive graph[J].Computer Engineering and Applications,2010,46(34):132-135,151.
[24]YAN R Q,ZHU Y S.ZHU Y S.Speech endpoint detection method based on signal recursion analysis[J].Journal on Communications,2007,23(4):35-39.
[25]HE Y,HE P Y,WANG S S.Method of speech enhancement based on recursive averaging and spectral subtraction[J].Computer Engineering & Applications,2009,45(8):221-223.
[26]ECKMANN J P,KAMPHORST S O,RUELLE D.Recurrence plots of dynamical systems[J].Europhysics Letter,1987,4(9):973-977.
[27]GUO X M,LI C P,LU D L.Application of quantitative recursive analysis and approximate entropy in heart sound classification and recognition[J].Journal of Data Acquisition & Processing,2013,28(5):559-564.
[28]TAKENS F.Detecting strange attractors in turbulence[J].Lecture Notes in Mathematics,1981,898(1):366-381.
[29]GAO J B,CAI H Q.On the structures and quantification of recurrence plots[J].Physics Letter A,2000,270(1／2):75-87.
[30]ZBILUT J P,WEBBER C L J.Embeddings and delays as derived from quantification of recurrence plots[J].Physics Letters A,1992,171:199-203.
[31]LI C L,YE N,HUANG H P,et al.Physiological signals emotion recognition based on recursive quantitative analysis[J].ComputerTechnology and Development,2018,28(11):94-98,102.
[32]ZHANG S C,LI X L,ZONG M.Efficient kNN Classification With Different Numbers of Nearest Neighbors[J].IEEE Transactions on Neural Networks and Learning Systems,2018,5(29):1-12.
[33]GANESH M A B,RATNADEEP M A B.Automatic Speech Recognition and Verifi-cation using LPC,MFCC and SVM[J].International Journal of Computer Applications,2015,127(8):47-52.
[34]DESHMUKH J,BHOSLE U.A study of mammogram classification using AdaBoost with decision tree,KNN,SVM and hybrid SVM-KNN as component classifiers[J].Journal of Information Hiding and Multimedia Signal Processing,2018,9(3):548-557.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

基于递归图分析的腭裂语音鼻漏气自动识别算法

Automatic Detection Algorithm of Nasal Leak in Cleft Palate Speech Based on Recursive Plot Analysis

PDF (PC)

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 2

Metrics

本文评价

推荐阅读 0

[1]	刘标,兰少华,张晶,刘光杰. 一种基于递归图的网络时间隐蔽信道检测方法 Approach Based on Recurrence Plot to Detect Covert Timing Channels 计算机科学, 2015, 42(2): 114-117. https://doi.org/10.11896/j.issn.1002-137X.2015.02.024
[2]	谢胜军,殷锋,周绪川. 基于递归率REC特征的网络流量相空间重构监测 Testing of Network Traffic Series in Reconstructed Phase Space Based on Recurrence Rate Feature 计算机科学, 2013, 40(11): 48-51.