| BackgroundSpasmodic dysphonia (SD) is one of relatively uncommon diseases in otorhinolaryngologydepartment. It is a disease of laryngeal muscle movement disorder, also known as focal laryngeal muscle dystonia. Currently, there are three types of SD, namely adductor SD, abductor SD and mixed SD, among which adductor SD is relatively common and accounts for more than 80%, but abductor SD and mixed SD have a low incidence. SD occurs in 30-50 years young and middle-aged people, among whom female patients account for about 60-85%. The overall incidence of SD is about 1/100,000, and its incidence shows a rising trend with the increase of knowledge and research on SD. The characteristic of adductor SD is vocal spasm during speech, which manifests as strained voice, voice tremor, strangled voice and frequent voice interruption; while the characteristic of abductor SD is that insufficient loudness leads to whispery and breathy voice or no voice. At present, no diagnostic criteria for SD are available in clinical practice, as well as corresponding specific examination methods. Most scholars hold the opinion that SD is a chronic nervous system disease caused by central motor nervous system disorder, which can be diagnosed and assessed mainly according to the following aspects:1. histories such as viral infection, life and work stress; 2. characteristics of voice such as involuntary sudden interruption, tremor and changes in rhythm and fluency; 3. morphology and movement of vocal cord in laryngoscopy; 4. laryngeal electromyography; 5. maximum phonation time and expiratory flow rate in laryngeal aerodynamic test; 6. sound wave graph; 7. the number and time of phonatory break during continuous phonation process in subjective auditory assessment, and meanwhile central and local pharyngolaryngeal organic diseases, neuromuscular paralysis diseases, mental disorders and other diseases should be excluded. Common assessment methods include voice acoustic analysis, subjective voice assessment, electroglottography analysis, sonograph and 3D sonagram analysis. Although voice acoustic analysis is relatively reliable, but it is also the most complex, and its clinical application is not convenient. In addition, no studies on the correlation between each assessment method have been reported, and subjective assessment lacks uniform dedicated phrases. Therefore, a simple and effective assessment method is required to improve the assessment ability, and promote the datamation and objectification of voice assessment. With adductor SD patients who speak mandarin Chinese as subjects, the present study aimed to assess and grade the patients through self-assessment of VHI, continuous phonation of vowel /a:/, reading proven stimuli phrases of SD, subjective assessment of acoustic judges, maximum phonation time (MPT), maximum loudness phonation time (MLPT) and computer voice analysis, and investigate whether there is correlation between subjective assessment and objective assessment through statistical analysis, so as to find more simple and effective assessment method and grading standard.Objective:With adductor SD patients who speak mandarin Chinese as subjects, the present study aimed to assess and grade the patients through self-assessment of VHI, continuous phonation of vowel /a:/, reading stimuli phrases of SD, subjective assessment of acoustic judges, MPT, MLPT and counting the percentage of phonatory break, number of frequency shift and percentage of aperiodic sound by computer voice analysis, and investigate whether there is correlation between subjective assessment and objective assessment through statistical analysis, so as to provide basis for formulating more simple and effective assessment method and grading standard.MethodsA total of 20 patients diagnosed as adductor SD in the otorhinolaryngologydepartment of Guangdong General Hospital from September 2014 to December 2015 were selected as SD group, with 3 males (median age:29 years; range:19-41 years) and 17 females (median age:32 years; range:21-59 years). The shortest course of disease was half year and the longest was 10 years, with a median of 3 years. A total of 20 healthy volunteers with corresponding age and gender with SD group were selected as normal control group, with 3 males (median age:30 years; range 22-44 years) and 17 females (median age:29 years; range:21-52 years).1. VHI score:All participants filled in Chinese version of Voice Handicap Index Scale, and the score was recorded.2. Measurement of MPT:Participants phonated /a:/ three times, and the time of the longest phonation was recorded.3. Measurement of MLPT:Participants phonated /a:/ three times with maximum loudness, and the time of the longest phonation was recorded.4. Voice assessment of acoustic judges:The voice assessment was performed by two acoustic judges under randomized and double-blind principle. (1) For continuous vowel, the voice was scored from breathy sound, rough sound and phonatory break three aspects:0-normal; 1-mild, difficult to be perceived; 2-moderate, can be perceived; 3-severe. (2) For stimuli phrases, the voice was scored from the number of phonatory break and fluency two aspects. After counting the total number of phonatory break in stimuli phrases, two acoustic judges calculated the mean value as the number of phonatory break. The fluency was scored according to the following standard:0-normal; 1-mild, there were hoarseness, tremor and breathy sound occasionally; 2-moderate, each sentence shifted the attention of acoustic judges and disturbed the fluency of words at least once; 3-severe, the semantic expression in communication was seriously affected. The mean value of two acoustic judges was calculated as the fluency.5. Voice analysis:The stimuli phrases were standard mandarin Chinese "ping guo,huo che,shi de, ni shi ge hao ren,wo men na bian you wang qiuchang, can guan jiu ba he yi ge mian bao dian,ta shi duo me kang kai a". Participants read the stimuli phrases with smooth speed and intonation, and the voice was recorded and analyzed using Praat5.0 software. (1) Number of phonatory break:Phonatory breaks that could be found in waveform included continuous vowel and read words. The number of phonatory break was counted manually, and the phonatory break should last for more than 50ms, with the phonatory break between word and syllable excluded. (2) Phonatory break time:The beginning and end of phonatory break was marked by line, with ms as unit. (3) Percentage of phonatory break time:The sum of phonatory break time in each word/total time of stimuli phrasesx100%. (4) The number of frequency shift:The number of frequency change above 50Hz within 50ms was counted manually, with first character excluded, the frequency shift data were divided into 2 types:50-100Hz and above 100Hz. (5) The number of aperiodic sound fragment was counted manually, including 3 types:0-49ms,50-100ms and above 100ms. The percentage of aperiodic sound was calculated:the sum of aperiodic sound fragment time in each word/total time of stimuli phrasesx100%.6. Statistical methods:1)Spearmem correlation analysis and Wilcoxon rank test were performed on the subjective acoustic assessment results of two acoustic judges to confirm the consistency and reliability of their subjective acoustic assessment results.2) Statistical analysis was performed using SPSS 19.0 software. Measurement data were expressed as mean ±SD, and enumeration data were expressed as frequency. Non-parametric rank sum test was used for inter-group comparison of abnormal distribution data, and paired t test was used for inter-group comparison of normal distribution data; Pearson correlation analysis was used to analyze the correlation of normal distribution data, and Spearman correlation analysis was used to analyze the correlation of abnormal distribution data. p< 0.05 was significant level.Results1. VHI:The VHI score of SD group was 87.90±19.62. The VHI score of normal control group was 3.35±5.69. Comparison of two groups showed that the VHI score of SD group was higher than that of normal control group, and the difference was statistically significant (p<0.01).2. MPT and MLPT:The MPT of SD group was 15.24±8.19, that of normal control group was 17.98±6.71, and there was no statistically significant difference between two groups (p>0.05); the MLPT of SD group was 12.15±4.68, that of normal control group was 19.18±5.17, thus the MLPT of SD group was significantly shorter than that of normal control group, and the difference was statistically significant (p<0.01).3. Subjective acoustic assessment results:1) Continuous vowel /a:/: The score of 20 adductor SD patients was2.05±0.71,among whom there were 5 mild cases,11 moderate cases and 4 severe cases. Voice tremor and rough sound that were difficult to be perceived existed in some individuals of the normal control group. The score of 20 individuals in normal control group was 0.30±0.34, among whom there were 18 normal cases and 2 mild cases. The overall impression score of SD patients on continuous vowel was significantly higher than that of normal control group (p<0.01).Reliability test showed that there was significant correlation in the judgment of continuous vowel/a:/ between two acoustic judges (r=0.868,p<0.01), and difference test showed that there was no significant difference between two judges (Z=-0.771,p>0.05), which suggested that the subjective assessment of continuous vowel /a:/ had high consistency.2) Stimuli phrases:(1) The number of phonatory break:The number of phonatory break in SD group was 4.68±4.38, with the minimum of 1 and the maximum of 19, while the number of phonatory break in normal control group was 0.00±0.00, and thus the number of phonatory break in SD group was significantly higher than that in normal control group (Z=-5.796,p<0.01). Reliability test showed that there was significant correlation in the judgment of the number of phonatory break between two acoustic judges (r=0.997,p<0.01), and difference test showed that there was no significant difference between two judges (Z=-0.447, p>0.05), which suggested that the subjective assessment of the number of phonatory break had high consistency.3) Fluency grade:The fluency score of stimuli phrases in SD group was 1.48±0.80, with 11 mild cases,6 moderate cases and 3 severe cases; the score in normal control group was 0.00±0.00, and all the cases were normal. The fluency score of stimuli phrases in SD group was significantly higher than that in normal control group, and the difference was statistically significant (Z=-5.611, p<0.01). Reliability test showed that there was significant correlation in the judgment of the fluency of reading stimuli phrases between two acoustic judges (r= 0.964, p<0.01), and difference test showed that there was no significant difference between two judges (Z=-1.732,p>0.05), which suggested that the subjective assessment of the fluency of reading stimuli phrases had high consistency.4. Voice analysis results1) Frequency shift was observed in both SD group and normal control group. In SD group, all the 20 patients presented frequency shift when reading stimuli phrases, while in normal control group,4 individuals did not present frequency shift when reading stimuli phrases. SD group had large number of frequency shift and high proportion of>100Hz type, and one syllable could present multiple types of frequency shift; normal control group had small number of frequency shift and high proportion of 50-100Hz type. The number of frequency shift in SD group was 27.00±14.12, while the number of frequency shift in normal control group was 6.15±5.17.2) In SD group, all the 20 patients presented aperiodic sound when reading stimuli phrases, while in normal control group,3 individuals did not present aperiodic sound. SD group had high frequency of aperiodic sound and large proportion of 50-100ms and> 100ms types, and one syllable could present multiple types; normal control group had low frequency and large proportion of 0-49ms type. The percentage of aperiodic sound in SD group was 7.25±5.22%, while the percentage of aperiodic sound in normal control group was 0.83±0.72%.3) All the individuals of normal control group did not present phonatory break when reading stimuli phrases, while SD patients usually presented phonatory break during conversion of consonant and vowel in a syllable or there might be loss of consonant in a syllable, and some patients presented phonatory break in the vowel of a syllable. The percentage of phonatory break in SD group was 1.77±3.24, while the percentage of phonatory break in normal control group was 0.00±0.00%.4) Comparison on the percentage of phonatory break, the percentage of aperiodic sound and the number of frequency shift between two groups by paired t test showed that the percentage of phonatory break, the percentage of aperiodic sound and the number of frequency shift in SD group were significantly higher than those in normal control group, and the difference was statistically significant (all p<0.05).5. Correlation analysis on each measure of SD group:Pearson correlation analysis was performed to investigate the correlation between each objective measure of SD group, and the result showed that the percentage of phonatory break in stimuli phrases had significant positive correlation with the percentage of aperiodic sound (r=0.606,p<0.01) and the number of phonatory break in stimuli phrases (r=0.849, p<0.01); the percentage of aperiodic sound had significant positive correlation with the number of phonatory break in stimuli phrases (r=0.581,p<0.01); MPT had significant positive correlation with MLPT (r=0.645,p<0.01).The fluency of reading stimuli phrases had significant positive correlation with the percentage of phonatory break (r=0.597,p<0.01), the percentage of aperiodic sound (r=0.491,p<0.05), the overall impression on continuous vowel/a:/(r=0.514,p <0.05) and the number of phonatory break in stimuli phrases (r=0.734,p<0.01).VHI had no correlation with MPT, MLPT, the percentage of aperiodic sound, the number of frequency shift, the overall impression on continuous vowel/a:/, the percentage of phonatory break in stimuli phrases, the number of phonatory break in stimuli phrases and the fluency of stimuli phrases.The number of frequency shift had no correlation with other subjective and objective measures.Conclusion1. The present study further demonstrated that SD stimuli phrases "ping guo, huo che,shi de, ni shi ge hao ren,wo men na bian you wang qiu chang, can guan, jiu ba he yi ge mian bao dian, ta shi duo me kang kai a" can well highlight the characteristic phonatory break, frequency shift and aperiodic sound in mandarin population.2. The vowel RBH score, phonatory break and fluency of ADSD patients in subjective acoustic assessment by acoustic judges are significantly worse than those of normal people, and the VHI of patients in subjective voice assessment is also worse than that of normal people.3. ADSD patients have different degrees of frequency shift, phonatory break and aperiodic sound in computer voice analysis, their quality of voice is significantly worse than that of normal people, and their MLPT is also significantly lower than that of normal people.4. Some subjective and objective measures of ADSD are correlated, and relatively simpler subjective assessment and MLPT can be used to replace the frequency shift, phonatory break and aperiodic sound tests in computer voice analysis.5. VHI of patients as a subjective voice assessment has limited consistency with objective examination, but it can be used as a reference method for ADSD assessment. |
PDF Full Text Request