Acoustic analysis of an articulation assessment and training system / 中华物理医学与康复杂志

Objective:To compare the differences in important parameters between the articulation assessment and training system of intelligently extracted speech with those from the Praat acoustic software and those manually extracted.Methods:The speech of thirty-two normal subjects was captured using the intelligent articulation assessment and training system and using Praat acoustic software. The former analyzed the mean fundamental frequencies (mF0s), the first formant peaks (F1s) and the second formant peak (F2s) of the sustained vowels /a/, /i/ and /u/. The speech parameters collected by the traditional Praat software were extracted and analyzed by professionals. The two tools′ consistency in terms of these important acoustic parameters was analyzed.Results:The results with all 32 subjects when retested returned ICC values above 0.9 with all three vowels with the exception of mF0 for /u/ (ICC=0.75), indicating excellent retest reliability for the articulation assessment and training system. The ICC values also indicated excellent consistency between the two kinds of software in analyzing mF0, F1 and F2 of the three vowels. The mF0, F1, F2, FCR, VAI, tongue spacing, VSA, and mandibular spacing of all three vowels were mostly distributed within the 95% confidence interval of the data points in Bland-Altman plots, indicating the high accuracy of both acoustic analysis systems in speech measurement. The mean fundamental frequency values of the male long vowels /a/, /i/ and /u/ were all significantly lower than for the female long versions.Conclusions:The retest reliability of the articulation assessment and training system was good, and the results of the articulation check in the natural state were in good consistency compared to the Praat check and were interchangeable in the articulation check.

Comparison of Voice Outcomes between Medialization Thyroplasty and Arytenoid Adduction with or without Injection Laryngoplasty in Unilateral Vocal Fold Paralysis Patients

BACKGROUND AND OBJECTIVES: In unilateral vocal fold paralysis (VFP), medialization thyroplasty (MT), arytenoid adduction (AA) and injection layrngoplasty (IL) are the most common procedures to correct phonatory problems. There is no consensus that which procedure is superior to the other for correcting the glottal insufficiency. The purpose of this study was to compare the phonatory parameters between MT, AA and AA with IL (AA+IL) in patients with unilateral VFP. MATERIALS AND METHODS: This retrospective study enrolled patients from 2005 to 2016. Total 72 patients (49 male, 23 female, mean age 54.5 years) were classified into three groups ; MT (n=28), AA (n=12), and AA+IL (n=32). GRBAS scales, maximum phonation time (MPT), jitter, shimmer, noise to harmonic ratio (NHR), and voice handicap index (VHI)-10 and VHI-30 were preoperatively and postoperatively collected and compared between the three groups. RESULTS: Age, gender and cause of VFP were not significantly different between the three groups. In MT and AA groups, MPT, VHI, G (overall grade) and B (breathiness) were significantly improved. In AA+IL group, jitter, shimmer, NHR, MPT, VHI, G and B were significantly improved. In analysis of differences (pre-postoperative values), Δ jitter (p < 0.001), Δ shimmer (p=0.031), and Δ NHR (p=0.002) were significantly different and AA+IL group showed the greatest improvement. CONCLUSION: Analysis of voice parameters showed that all the three procedures for patients with unilateral VFP are effective in the improvement of voice ; especially in MPT, VHI-10, G and B scales. Compared to the others, AA+IL provided the better acoustic values including jitter, shimmer and NHR.

Change of Voice Quality before and after Treatment of Short-Term Therapy with Proton Pump Inhibitor in Laryngopharyngeal Reflux / 대한이비인후과학회지

BACKGROUND AND OBJECTIVES: The aim of this study was to verify that voice analysis could be an alternate tool for the evaluation of proton pump inhibitor treatment of laryngopharyngeal reflux (LPR). SUBJECTS AND METHOD: Twenty-two patients with LPR symptoms underwent laryngoscopy, stroboscopy and their reflux finding index (RFI) were evaluated. Subjective reflux symptom scores (RSS) and voice handicap index (VHI) were completed at the baseline. All patients underwent voice analysis. Thereafter, patients had short-term proton pump inhibitor therapy for 6 weeks. The RFI, RSS, VHI and voice analysis were repeated during the last week of the treatment. RESULTS: RFI was improved and both RSS and VHI were improved after the treatment. RSS and VHI were significantly correlated. In voice analysis, shimmer significantly improved and harmony to noise ratio (HNR) also improved. CONCLUSION: Voice analysis can be indicators of treatment results of laryngopharyngeal reflux disease.

The Effect on Voice by Strap Muscle Cutting in Thyroidectomy / 대한내분비외과학회지

PURPOSE: Voice change after thyroidectomy has generally been the result of damage to the recurrent or superior laryngeal nerve. But many patients complain voice alteration without laryngeal nerve injury after thyroidectomy. The purpose of this study is to investigate whether strap muscle division results in any subjective or objective functional sequelae in voice, through long-term follow-up prospectively. METHODS: Twenty-two female patients who had undergone thyroid surgery between July 1998 and December 1999, were studied. The patients who were planned for neck dissection, who had benign laryngeal disease or vocal cord paresis, and whose vocal cord paresis were developed after thyroid surgery, were excluded from this study. Twelve patients had undergone thyroidectomy via retraction of strap muscle and ten patients had undergone thyroidectomy via cutting of strap muscle. For evaluation of voice, questionnaires for changes of voice, acoustics (fundamental frequency, jitter, shimmer, signal to noise ratio, noise to harmonic ratio, voice range), and aerodynamic (maximal phonation time) analyses were done. RESULTS: The subjective voice symptoms after thyroidectomy were disturbances of high pitch, singing, loud voice, and easy fatigue at phonation. There were no significant differences in voice parameters on acoustic and aerodynamic analyses between the strap muscle retraction group and the cutting group through long-term follow-up. CONCLUSION: We conclude that strap muscle division does not result in any subjective or objective functional problems in voice. We suggest that surgical division and reconstruction of these muscles should be employed routinely when operating on large, toxic or neoplastic glands.

Evaluation of the Voice Parameters Following Supracricoid Partial Laryngectomy / 대한이비인후과학회지

BACKGROUND AND OBJECTIVES: Conservation laryngeal surgery is designed to remove the cancer mass completely while preserving the physiologic functions of the larynx. Recently, the supracricoid partial laryngectomy (SCPL) has broadened the spectrum of reliable techniques available to conservation laryngeal surgeons. Much research has been devoted to the surgical technique and oncological results of SCPL, but only a few signi6cant objective reports of the phonatory results are available presently. This study was designed to analyze the phonatory results achieved after SCPL for laryngeal cancer. MATERIALS AND METHODS: Selected characteristics of the speech and voice were compared in 25 patients who had undergone SCPL and 10 normal adult laryngeal (NAL) speakers. Durational and frequency features were analyzed with the CSL and the MDVP. RESULTS: SCPL speech proved comparable to NAL speech at the average fundamental frequency. SCPL spech and voice were statistically less efficient than NAL speech in jitter, shimmer, harmonics-to-noise ratio, mean airflow rate, maximal phonation time, and subglottic pressure (p<0.01). The completion of an arytenoid cartilage resection and/or anterior cricoid cartilage removal did not statistically modify the durational and frequency features of SCPL speakers (p <0.01). CONCLUSION: The voice parameters of SCPL speakers were different from those of the NAL speakers, but patients were thought to be allowed social interaction.