Analysis Of The Voice Characteristics Of Patients With Obstructive Sleep Apnea

Based on the voice signals of domestic snoring patients,this paper investigates the differences in voice acoustic parameters and electroglottograph parameters when pronouncing the five vowels /a/,/o/,/e/,/i/,/u/ in Mandarin in a seated position to find objective assessment indicators that can be used as a large-scale aid in screening patients with obstructive sleep apnea(OSA).The physiological parameters characterizing OSA were included for correlation analysis to deepen the etiological analysis of voice disorders in OSA patients and to provide a basis for the treatment of OSA patients and their voice disorders.The 138 adult OSA patients diagnosed by polysomnography(PSG)were grouped according to the different degrees of apnea hypopnea index(AHI).125 of them were OSA patients and were classified as the experimental group;13 were simple snoring(SS)patients and were classified as the control group.The example words were pronounced at a natural and comfortable volume to unitary rhyme with the Mandarin Yinping(First tone)according to the cue.Voice acoustic parameters were collected using a super nasal-oral ratiometry system(SNORS)and voice physiological parameters were collected using electroglottography(EGG).The Praat acoustic analysis software was run to extract the acoustic parameters of the unitary rhymes in the example words,such as fundamental frequency perturbation(jitter),amplitude perturbation(shimmer),and harmonics-to-noise ratio(HNR);The EGGWorks voice analysis software was run to extract parameters from the electroglottography of the unitary rhymes,such as contact quotient perturbation(CQP),contact quotient(CQ),contact index perturbation(CIP),contact index(CI),and speed quotient(SQ).It was found that there were statistical differences(p<0.05)between the SS and OSA groups when pronouncing HNR for the low vowel /a/,shimmer,CIP,and SQ for the high vowel /i/,and shimmer,HNR,and SQ for the high vowel /u/,and no statistical differences(p>0.05)in all parameters when pronouncing the Mandarin semi-high vowels /o/ and /e/.It was suggested that Mandarin low vowels,semi-high vowels and high vowels differed in their sensitivity to reflect vocal cold vibration and glottic closure in SS and OSA patients,with high vowels being the most effective,low vowels the second most effective and semi-high vowels the least effective.The best clinically defined cut-off values were obtained by binary logistic regression analysis and ROC curve analysis,and HNR for low vowel/a/,shimmer and HNR for high vowel/u/ had diagnostic value for screening OSA patients,with shimmer for high vowel/u/ having the largest area under the ROC curve by statistical test and being the most suitable as an aid to screen OSA patients objective assessment index.The relevant physiological parameters characterizing OSA were included in the study,and it was found that: the lateral pharyngeal wall stenosis in SS patients was not optimistic as in OSA patients in the subjective assessment parameters(p>0.05),and the degree of tongue root grade was lower(p<0.05);there was no statistical difference between the two groups in age and height in the objective assessment parameters(p>0.05),and there was a statistical difference in weight,BMI,neck circumference,abdominal circumference,hip circumference,and waist-to-hip ratio(p<0.05).In the correlation analysis between AHI and physiological parameters of OSA patients,the objective clinical parameters related to obesity had better results,suggesting that obesity would lead to the aggravation of OSA patients’ condition;in the correlation analysis with voice parameters,the electroglottograph parameters had better results,suggesting that the electroglottograph detection method was more realistic and accurate compared with voice acoustic detection.Overall,the correlation between AHI and physiological parameters was stronger than that between AHI and voice parameters in OSA patients,suggesting that physiological parameters characterizing OSA have an important influence.Further correlation analysis of voice parameters and physiological parameters in OSA patients revealed that age,height,waist-to-hip ratio and subjective assessment parameters were poorly tested for correlation with voice disorders in OSA patients,and studying voice problems in OSA patients requires more attention to objective clinical parameters related to obesity.When OSA patients are in the overweight or obese range,the soft tissues around the upper airway become fatty and hypertrophic during sleep,which makes it easier for them to fall back and block the upper airway,resulting in mouth breathing,snoring and recurrent sleep apnea symptoms,which aggravate voice disorders.