Evaldas Padervinskis

Validation of the Acoustic Voice Quality Index in the Lithuanian Language

OBJECTIVES The aim of the present study was to validate the Acoustic Voice Quality Index in Lithuanian language (AVQI-LT) and investigate the feasibility and robustness of its diagnostic accuracy, differentiating normal and dysphonic voice. METHODS A total of 184 native Lithuanian subjects with normal voices (n = 46) and with various voice disorders (n = 138) were asked to read aloud the Lithuanian text and to sustain the vowel /a/. A sentence with 13 syllables and a 3-second midvowel portion of the sustained vowel were edited. Both speech tasks were concatenated, and perceptually rated for dysphonia severity by five voice clinicians. They rated the Grade (G) from the Grade Roughness Breathiness Asthenia Strain (GRBAS) protocol and the overall severity from the Consensus Auditory-perceptual Evaluation of Voice protocol with a visual analog scale (VAS). The average scores (Gmean and VASmean) were taken as the perceptual dysphonia severity level for every voice sample. All concatenated voice samples were acoustically analyzed to receive an AVQI-LT score. RESULTS Both auditory-perceptual judgment procedures showed sufficient strength of agreement between five raters. The results achieved significant and marked concurrent validity between both auditory-perceptual judgment procedures and AVQI-LT. The diagnostic accuracy of AVQI-LT showed for both auditory-perceptual judgment procedures comparable results with two different AVQI-LT thresholds. The AVQI-LT threshold of 2.97 for the Gmean rating obtained reasonable sensitivity = 0.838 and excellent specificity = 0.937. For the VAS rating, an AVQI-LT threshold of 3.48 was determined with sensitivity = 0.840 and specificity = 0.922. CONCLUSIONS The AVQI-LT is considered a valid and reliable tool for assessing the dysphonia severity level in Lithuanian-speaking population.

Reirradiation for patients with recurrence head and neck squamous cell carcinoma: A single-institution comparative study

BACKGROUND AND OBJECTIVE In the last decade, the number of publications that report on the use of external beam radiotherapy and high-dose-rate brachytherapy (HDR-BRT) in the treatment of recurrent head and neck cancer has increased, but no studies compare external beam radiotherapy and HDR-BRT. The aim of this study was to evaluate and to compare the efficacy and toxicity of the three-dimensional conformal radiotherapy (3D-CRT) and HDR-BRT in the treatment of recurrent head and neck cancer. MATERIAL AND METHODS A total of 64 patients with head and neck cancer recurrence were randomly assigned at a 1:1 ratio to receive either 3D-CRT (50Gy/25 fractions) in the control group or HDR-BRT (30Gy/12 fraction) in the experimental group. RESULTS The overall survival rate of patients treated with HDR-BRT at 1 and 2-years was 74% and 67%, respectively, compare to 3D-CRT group - 51% and 32%, respectively (P=0.002). Local control at 1- and 2-years in patients who received HDR-BRT was 77% and 63% compare with 47% and 25%, respectively, for the patients who received the 3D-CRT (P<0.001). Most patients developed mild to moderate acute mucositis and dermatitis. In the 3D-CRT group, severe late toxicity was determined in 11 patients (35.5%), and in the HDR-BRT group, in 1 patient (3.1%) (P=0.001). There was no grade 5 toxicity. CONCLUSIONS Following our results, we concluded that HDR-BRT is a more effective and safer treatment approach for head and neck cancer recurrences than 3D-CRT.

Data dependent random forest applied to screening for laryngeal disorders through analysis of sustained phonation: acoustic versus contact microphone.

Comprehensive evaluation of results obtained using acoustic and contact microphones in screening for laryngeal disorders through analysis of sustained phonation is the main objective of this study. Aiming to obtain a versatile characterization of voice samples recorded using microphones of both types, 14 different sets of features are extracted and used to build an accurate classifier to distinguish between normal and pathological cases. We propose a new, data dependent random forests-based, way to combine information available from the different feature sets. An approach to exploring data and decisions made by a random forest is also presented. Experimental investigations using a mixed gender database of 273 subjects have shown that the perceptual linear predictive cepstral coefficients (PLPCC) was the best feature set for both microphones. However, the linear predictive coefficients (LPC) and linear predictive cosine transform coefficients (LPCTC) exhibited good performance in the acoustic microphone case only. Models designed using the acoustic microphone data significantly outperformed the ones built using data recorded by the contact microphone. The contact microphone did not bring any additional information useful for the classification. The proposed data dependent random forest significantly outperformed the traditional random forest.

Combined Use of Standard and Throat Microphones for Measurement of Acoustic Voice Parameters and Voice Categorization

OBJECTIVE The aim of the present study was to evaluate the reliability of the measurements of acoustic voice parameters obtained simultaneously using oral and contact (throat) microphones and to investigate utility of combined use of these microphones for voice categorization. MATERIALS AND METHODS Voice samples of sustained vowel /a/ obtained from 157 subjects (105 healthy and 52 pathological voices) were recorded in a soundproof booth simultaneously through two microphones: oral AKG Perception 220 microphone (AKG Acoustics, Vienna, Austria) and contact (throat) Triumph PC microphone (Clearer Communications, Inc, Burnaby, Canada) placed on the lamina of thyroid cartilage. Acoustic voice signal data were measured for fundamental frequency, percent of jitter and shimmer, normalized noise energy, signal-to-noise ratio, and harmonic-to-noise ratio using Dr. Speech software (Tiger Electronics, Seattle, WA). RESULTS The correlations of acoustic voice parameters in vocal performance were statistically significant and strong (r = 0.71-1.0) for the entire functional measurements obtained for the two microphones. When classifying into healthy-pathological voice classes, the oral-shimmer revealed the correct classification rate (CCR) of 75.2% and the throat-jitter revealed CCR of 70.7%. However, combination of both throat and oral microphones allowed identifying a set of three voice parameters: throat-signal-to-noise ratio, oral-shimmer, and oral-normalized noise energy, which provided the CCR of 80.3%. CONCLUSIONS The measurements of acoustic voice parameters using a combination of oral and throat microphones showed to be reliable in clinical settings and demonstrated high CCRs when distinguishing the healthy and pathological voice patient groups. Our study validates the suitability of the throat microphone signal for the task of automatic voice analysis for the purpose of voice screening.

Fusing voice and query data for non-invasive detection of laryngeal disorders

Voice and query data are explored for the task of laryngeal disorders detection.Decision-level fusion by complete-case analysis is compared to imputation strategies.Query data outperform voice, fusion after iterative model-based imputation - the best.Human readable rules were extracted from the query data using affinity analysis. Topic of this study is exploration and fusion of non-invasive measurements for an accurate detection of pathological larynx. Measurements for human subject encompass answers to items of a specific survey and information extracted by the openSMILE toolkit from several audio recordings of sustained phonation (vowel /a/). Clinical diagnosis, assigned by medical specialist, is a target attribute distinguishing subject as healthy or pathological. Random forest (RF) is used here as a base-learner and also as a meta-learner for decision-level fusion. 5 RF classifiers, built separately on 3 variants of audio recording data (raw and after two types of voice activity detection) and 2 variants of questionnaire (with 9 and 26 questions) data, are fused selectively by finding out the best combination of all possible. Before fusion, due to presence of missing values in query modalities, several imputation techniques were evaluated besides the complete-case analysis by listwise deletion. Out-of-bag equal error rate (EER) was found to be higher for audio data and lower for query, but each variant was outperformed by the decision-level fusion. Fusion after listwise deletion provided EER of 4.84%, meanwhile imputation was found to improve detection slightly and helped to achieve EER of 4.55%. Variable importance, as permutation-based mean decrease in RF accuracy, was reported for query and audio data. Finally, regarding the noteworthy performance of the query data, 22 association rules (11 healthy + 11 pathological) using 17 questions were obtained for comprehensible insights.

Fusing Various Audio Feature Sets for Detection of Parkinson’s Disease from Sustained Voice and Speech Recordings

The aim of this study is the analysis of voice and speech recordings for the task of Parkinson’s disease detection. Voice modality corresponds to sustained phonation /a/ and speech modality to a short sentence in Lithuanian language. Diverse information from recordings is extracted by 22 well-known audio feature sets. Random forest is used as a learner, both for individual feature sets and for decision-level fusion. Essentia descriptors were found as the best individual feature set, achieving equal error rate of 16.3 % for voice and 13.3 % for speech. Fusion of feature sets and modalities improved detection and achieved equal error rate of 10.8 %. Variable importance in fusion revealed speech modality as more important than voice.

A Rare Case of Large Schwannoma of the Parapharyngeal Space

Schwannoma originating from the peripheral nerves is a rare lesion of the parapharyngeal space. The special traits of the presented case included the following: the patient presented with slowly progressing dysphagia, speech difficulties, jaw numbness, and taste dysfunction. A dislocated lateral pharyngeal wall with mild inflammatory changes of the oropharyngeal mucosa was observed during pharyngoscopy. The radiological and histological characteristics of the neoplasm are consequently presented. Special emphasis is placed on the surgical treatment of the tumor.

Comparison of Sino-Nasal Outcome Test 22 Symptom Scores in Rhinogenic and Odontogenic Sinusitis

Background Odontogenic maxillary sinusitis (OMS) and rhinogenic sinusitis (RS) are the main types of chronic rhinosinusitis (CRS) and have a significant impact on health-related quality of life (HRQL), but the difference in HRQL and symptom presentation between them has not been specifically evaluated to date. Obejctive: Our aim was to compare patterns of symptoms and HRQL disease-specific domains in patients affected with these 2 types of CRS. Methods A group of 201 patients with CRS (99 with rhinogenic and 102 with odontogenic origin) completed the Sino-Nasal Outcome Test 22 (SNOT-22) questionnaire before treatment. Data sets were analyzed by using principal component analysis (PCA) to identify a set of symptom components together with the items excluded from PCA, which were then analyzed for differences between patients with OMS and RS. Results PCA of SNOT-22 items identified 5 components: “rhinologic,” “extranasal rhinologic,” “ear/facial,” “sleep and functional disturbance,” and “emotional disturbance.” Sneezing was excluded from PCA and treated as separate outcome variable and was significantly worse in RS patients. Patients with OMS scored significantly higher scores with regard to emotional disturbance, while RS patients scored significantly worse in sleep and functional disturbance. The extra symptom “malodor” was the most different symptom and was significantly worse in OMS patients. The total SNOT-22 score was not significantly different between the groups. Conclusion With controlling of covariates that may influence the severity of the disease, this study showed some significant differences in symptom patterns and HRQL impairment between patients with OMS and RS. Malodor is the most characteristic feature of OMS. Therefore, OMS should always be suspected in patients complaining of bad breath.

Exploring sustained phonation recorded with acoustic and contact microphones to screen for laryngeal disorders

Exploration of various features and different structures of data dependent random forests in screening for laryngeal disorders through analysis of sustained phonation recorded by acoustic and contact microphones is the main objective of this study. To obtain a versatile characterization of voice samples, 14 different sets of features were extracted and used to build an accurate classifier to distinguish between normal and pathological cases. We proposed a new, data dependent random forest-based, way to combine information available from the different feature sets. An approach to exploring data and decisions made by a random forest was also presented. Experimental investigations using a mixed gender database of 273 subjects have shown that the Perceptual linear predictive cepstral coefficients (PLPCC) was the best feature set for both microphones. However, the LP-coefficients and LPCT-coefficients feature sets exhibited good performance in the acoustic microphone case only. Models designed using the acoustic microphone data significantly outperformed the ones built using data recorded by the contact microphone. The contact microphone did not bring any additional information useful for classification. The proposed data dependent random forest significantly outperformed traditional designs.