Voice changes in patients with chronic obstructive pulmonary disease

Exploring explainable AI features in the vocal biomarkers of lung disease

This review delves into the burgeoning field of explainable artificial intelligence (XAI) in the detection and analysis of lung diseases through vocal biomarkers. Lung diseases, often elusive in their early stages, pose a significant public health challenge. Recent advancements in AI have ushered in innovative methods for early detection, yet the black-box nature of many AI models limits their clinical applicability. XAI emerges as a pivotal tool, enhancing transparency and interpretability in AI-driven diagnostics. This review synthesizes current research on the application of XAI in analyzing vocal biomarkers for lung diseases, highlighting how these techniques elucidate the connections between specific vocal features and lung pathology. We critically examine the methodologies employed, the types of lung diseases studied, and the performance of various XAI models. The potential for XAI to aid in early detection, monitor disease progression, and personalize treatment strategies in pulmonary medicine is emphasized. Furthermore, this review identifies current challenges, including data heterogeneity and model generalizability, and proposes future directions for research. By offering a comprehensive analysis of explainable AI features in the context of lung disease detection, this review aims to bridge the gap between advanced computational approaches and clinical practice, paving the way for more transparent, reliable, and effective diagnostic tools.

A Scoping Review on the Intersection Between Voice and Swallowing Measures in Healthy and Disordered Populations

PURPOSE

Voice and swallowing are distinct functions that share anatomical and physiological properties; however, research investigating their intersection is limited. The purpose of this scoping review was to explore the literature surrounding the relationship between voice and swallowing measures in healthy adults and those with non-degenerative disorders. Specifically, we aimed to elucidate whether objective voice measures could be used as correlates of swallowing function.

METHOD

We systematically searched four databases (Embase, PubMed, CINAHL, and Web of Science) for relevant literature using a combination of key words and controlled vocabulary generated from the Yale Mesh Analyzer. The inclusion criteria consisted of peer-reviewed studies in the English language that reported on healthy adults and/or patients with non-degenerative neurological disorders and pulmonary diseases and contained instrumental and/or objective voice and swallowing measures. Two raters completed the abstract screening process followed by independent full-text reviews. Case studies, review studies, gray literature, or abstract-only studies were excluded.

RESULTS

Among 5,485 screened studies, 182 were fully reviewed, with only 11 studies meeting the inclusion criteria. Eight studies found an association between voice and swallowing objective measures, whereas the other three did not. Significant voice measures that were related to swallowing safety and/or physiology included maximum fundamental frequency (F0), F0 range, maximum phonation time, biomechanics of effortful pitch glides, and voice onset time.

CONCLUSIONS

Although there was heterogeneity in the measures used, specific objective voice measures showed promise in clinical practice as a screening tool for dysphagia. Further investigations are needed to validate the clinical utility of these measures across diverse patient populations.

Voice quality after surgery or radiotherapy for glottic T1 squamous cell carcinoma: Results of the VOQUAL study

PURPOSE

Many series have compared voice quality after radiotherapy or surgery for cT1 glottic carcinoma. Different meta-analyses identify better results for radiotherapy while others do not identify any difference, some finally find a superiority of surgery. The purpose of this study was to compare the voice quality in the long term of patients who underwent transoral surgery versus exclusive irradiation for the treatment of cT1 glottic carcinoma.

MATERIAL AND METHODS

The VOQUAL study was a pilot comparative multicenter cross-sectional study. The primary endpoint was the Voice Handicap Index comparison between two groups (radiotherapy or surgery). The voice assessment also consisted in the heteroevaluation of voice quality by the Grade, Roughness, Breathness, Asthenia, and Strain rating scale reported by Hirano.

RESULTS

The study included 41 adult patients with cT1 carcinoma of the vocal cord treated by cordectomy or exclusive radiation in two oncologic centers. The median Voice Handicap Index value was 20 [8; 32.5] in the surgery group and 10 [4; 18.5] in the radiotherapy group. There was no statistically significant difference between the median values and the various components F, P and E of the questionnaire (P=0.1585). The median value of the numeric dysphonia Grade, Roughness, Breathness, Asthenia, and Strain scale was 2 [0; 5] in the surgery group and 2 [0.25; 3.75] in the radiotherapy group. There was no statistically significant difference between these values (P=0.78).

CONCLUSION

Our study did not show any significant difference on the primary endpoints of Voice Handicap Index and Grade, Roughness, Breathness, Asthenia, and Strain scores.

LEVEL OF EVIDENCE

III.

CLINICAL TRIAL REGISTRATION

The VOQUAL study was registered on the ClinicalTrials.gov platform under the number NCT04447456, in July 2020.

Maximum Phonation Times as Biomarkers of Lung Function

PURPOSE

To verify whether measurements of maximal phonation times are biomarkers of forced vital capacity in patients with chronic obstructive pulmonary disease, and to characterize the vocal aspects of these patients, taking into account variables, such as age, body mass index, use of bronchodilators, presence of symptoms, and quality of life related to voice.

METHODS

Complete records of 25 subjects with chronic obstructive pulmonary disease, both sexes, aged 31 to 85 years, evaluated by forced vital capacity, maximum phonation times of /a/, and numerical count and number reached at this count, Vocal Symptom Scale, Voice Quality of Life. Data were presented descriptively and statistically analyzed using Student's t test for independent samples and Mann-Whitney U test. A significance level of 5% was accepted. The receiver operating characteristic curve was plotted and the standardized value of forced vital capacity <80% was considered as an indicator of pulmonary dysfunction.

RESULTS

Patients exhibited reduced maximum phonation times for /a/, numeric counting, and reached digits in counting; discrepancies in Vocal Signs and Symptoms and Voice Quality of Life Scale scores. Numeric counting times of up to 12.5 seconds indicated that forced vital capacity may be impaired.

CONCLUSION

The patients with chronic obstructive pulmonary disease examined in this study exhibited vocal deviations as evidenced by reduced maximum phonation times of /a/, numeric counting, and the digit reached during counting, as well as deviations in vocal self-assessment. Maximum phonation time in numerical counting was considered a biomarker of pulmonary function impairment.

An Experimental Analysis on Multicepstral Projection Representation Strategies for Dysphonia Detection

Biometrics-based authentication has become the most well-established form of user recognition in systems that demand a certain level of security. For example, the most commonplace social activities stand out, such as access to the work environment or to one's own bank account. Among all biometrics, voice receives special attention due to factors such as ease of collection, the low cost of reading devices, and the high quantity of literature and software packages available for use. However, these biometrics may have the ability to represent the individual impaired by the phenomenon known as dysphonia, which consists of a change in the sound signal due to some disease that acts on the vocal apparatus. As a consequence, for example, a user with the flu may not be properly authenticated by the recognition system. Therefore, it is important that automatic voice dysphonia detection techniques be developed. In this work, we propose a new framework based on the representation of the voice signal by the multiple projection of cepstral coefficients to promote the detection of dysphonic alterations in the voice through machine learning techniques. Most of the best-known cepstral coefficient extraction techniques in the literature are mapped and analyzed separately and together with measures related to the fundamental frequency of the voice signal, and its representation capacity is evaluated on three classifiers. Finally, the experiments on a subset of the Saarbruecken Voice Database prove the effectiveness of the proposed material in detecting the presence of dysphonia in the voice.

Multi-Modal Point-of-Care Diagnostics for COVID-19 Based on Acoustics and Symptoms

BACKGROUND

The COVID-19 pandemic has highlighted the need to invent alternative respiratory health diagnosis methodologies which provide improvement with respect to time, cost, physical distancing and detection performance. In this context, identifying acoustic bio-markers of respiratory diseases has received renewed interest.

OBJECTIVE

In this paper, we aim to design COVID-19 diagnostics based on analyzing the acoustics and symptoms data. Towards this, the data is composed of cough, breathing, and speech signals, and health symptoms record, collected using a web-application over a period of twenty months.

METHODS

We investigate the use of time-frequency features for acoustic signals and binary features for encoding different health symptoms. We experiment with use of classifiers like logistic regression, support vector machines and long-short term memory (LSTM) network models on the acoustic data, while decision tree models are proposed for the symptoms data.

RESULTS

We show that a multi-modal integration of inference from different acoustic signal categories and symptoms achieves an area-under-curve (AUC) of 96.3%, a statistically significant improvement when compared against any individual modality ([Formula: see text]). Experimentation with different feature representations suggests that the mel-spectrogram acoustic features performs relatively better across the three kinds of acoustic signals. Further, a score analysis with data recorded from newer SARS-CoV-2 variants highlights the generalization ability of the proposed diagnostic approach for COVID-19 detection.

CONCLUSION

The proposed method shows a promising direction for COVID-19 detection using a multi-modal dataset, while generalizing to new COVID variants.

Evaluation of medical adherence, adverse drug reactions, and quality of life in post tubercular obstructive airway disease

Background

Increasing incidence of tuberculosis is intensifying the posttubercular obstructive airway disease (PTOAD) in developing countries. Currently, there are no standard treatment guidelines for the management of PTOAD patients. The present study aims to evaluate the prescribing pattern, adherence, adverse drug reactions (ADRs), and quality of life (QoL) in PTOAD patients.

Materials and Methods

A prospective observational study was conducted to evaluate the prescriptions of PTOAD patients, estimating the medical adherence using Morisky 8-Item Medication Adherence Questionnaire, assessing ADRs using Hartwig's Severity Assessment Scale and assessing QoL using St. George's respiratory Questionnaire. Chi-square test, analysis of variance, paired t-test were used to compare the data. The significance of change in adherence status was assessed by Wilcoxon signed-rank test.

Results

A total of 94 prescriptions of PTOAD were analyzed. Inhaled long-acting muscarinic antagonist was prescribed to 31.9% of patients. The most common inhaled fixed dose combination was long-acting beta-2 agonist with corticosteroid, prescribed to 52.1% of patients. At final follow-up, maximum percentage of patients were found to be highly adherent, i.e. 56.4%. Overall, 34% of patients have complained about mild category of ADRs. A significant improvement in QoL was observed. At baseline, mean forced expiratory volume in 1 (FEV1) was 64.66% ±23.61%, which increased significantly to 73.34% ±21.60% on final follow-up (P < 0.001).

Conclusion

Bronchodilators are the mainstay of treatment of PTOAD patients, since both the QoL and FEV1 were improved with treatments. However, to have good treatment outcome, strict adherence along with safety of the medications must be assured.

Voice Problems and Related Risk Factors Among Hairdressers

OBJECTIVE

Hairdressers are more prone to respiratory diseases, bronchitis, asthma, shortness of breath, and cough due to the nature of their occupation and the constant use of cosmetics. Therefore, they may be prone to voice disorders or laryngeal problems. Voice problems may affect not only their social and emotional relationships but also their jobs. The current study was conducted to investigate voice problems and related risk factors in hairdressers.

DESIGN

Cross-sectional, descriptive-analytical.

PATIENTS AND METHODS

A total number of 293 women participated in the study. The study group consisted of 147 hairdressers and the control group consisted of 146 women with other jobs. All study participants were requested to complete a self-reported questionnaire. This questionnaire investigated demographic characteristics, voice problems and symptoms, laryngeal discomfort, working features, and workplace conditions. Chi-square, independent t-test, Fisher's exact test, and logistic regression were used to analyze the data.

RESULTS

The prevalence of voice problems in hairdressers and the control group was 33.33% and 15.75%, respectively, and this difference was statistically significant (P < 0.001). Results of the Chi-Square test showed that prevalence of hoarseness, vocal fatigue, dryness, and cough were higher in hairdressers than the control group (P < 0.05). According to regression logistic results, work time (hours per week) (P = 0.014; OR = 2.35; CI = 1.18-4.66) and presence of phonotraumatic behaviors (P = 0.012; OR = 2.73; CI = 1.24-5.96) increased the possibility of increasing the presence of voice symptoms among hairdressers.

CONCLUSION

The findings revealed that self-reported voice problems were more prevalent in the hairdressers group than in the control group and therefore the hairdressers were more prone to an increased risk of developing voice problems. The most common symptoms in the hairdressers group were cough, hoarseness, and dryness in the vocal tract. Based on these results, in addition to paying attention to recommendations related to environmental issues and exposure to chemical hazards, hairdressers may benefit from receiving appropriate training in voice production, voice disorders, and the prevention of voice disorders.

Swallowing and laryngeal complications in lung and heart transplantation: Etiologies and diagnosis

Despite continued surgical advancements in the field of cardiothoracic transplantation, post-operative complications remain a burden for the patient and the multidisciplinary team. Lesser-known complications including swallowing disorders (dysphagia), and voice disorders (dysphonia), are now being reported. Such disorders are known to be associated with increased morbidity and mortality in other medical populations, however their etiology amongst the heart and lung transplant populations has received little attention in the literature. This paper explores the potential mechanisms of oropharyngeal dysphagia and dysphonia following transplantation and discusses optimal modalities of diagnostic evaluation and management. A greater understanding of the implications of swallowing and laryngeal dysfunction in the heart and lung transplant populations is important to expedite early diagnosis and management in order to optimize patient outcomes, minimize allograft injury and improve quality of life.

The Impact of Respiratory Function on Voice in Patients with Presbyphonia

BACKGROUND AND OBJECTIVE

Presbyphonia is an age-related voice disorder characterized by vocal fold atrophy and incomplete glottal closure during phonation. The extent to which the effects of presbyphonia may be compounded by age-related declines in the respiratory system and further impact communication and quality of life remains unknown. Therefore, the objective of this study was to determine how variations in respiratory function impacts voice measures in a sample of participants with presbyphonia.

METHODS

In this pilot study, 21 participants with presbyphonia underwent respiratory assessments (spirometry and respiratory muscle strength testing) and voice assessments (videostroboscopy, acoustic analysis, auditory-perceptual ratings, aerodynamic assessment, and self-assessments). Factor and cluster analyses were conducted to extract voice and respiratory constructs and to identify groups of participants with similar profiles. Correlations and regression analyses were conducted to better describe the relationships between voice and respiratory function.

RESULTS

Respiratory function was found to impact voice via two main pathways: through its physiological effect on voice and through its impact on general health and impairment. A lower respiratory function was associated with a lower vocal fold pliability and regularity of vibration and with an elevated aerodynamic resistance accompanied by laryngeal hyperfunction. Standardized measures of respiratory function were associated with perceived voice-related handicap. Respiratory function did not associate with voice quality, which was mostly influenced by the severity of vocal fold atrophy.

CONCLUSION

Poor respiratory health exacerbates the burden of vocal fold atrophy and, therefore, implementation of respiratory screening prior to starting voice therapy may significantly affect the treatment plan and consequently the outcomes of voice therapy in this patient population.

The Utility of Pulmonary Function Testing in Patients Presenting With Dysphonia

OBJECTIVE

We aimed to evaluate the utility of pulmonary function testing (PFT), particularly forced expiratory flow (FEF) 25-75%, in patients presenting with dysphonia.

STUDY DESIGN

A retrospective chart review was carried out.

METHODS

Records of 199 patients who presented with dysphonia were reviewed to determine whether in-office PFTs, which we perform routinely, lead to new pulmonary diagnoses or the need for additional pulmonary medications, after assessment by a pulmonologist. Of particular interest was evaluating if FEF25-75% of predicted values less than 80% can be used as a marker for occult pulmonary disease in patients presenting with dysphonia.

RESULTS

Of the 199 patient charts reviewed, 129 were female and 70 were male. The age of patients ranged from 18 to 88 years, with a mean of 46.8 years. The body mass index ranged from 17.5 to 53.4 kg/m². One hundred five (52.8%) patients had FEF25-75% values less than 80% of predicted (poor midflow values). Of these patients, 76 (72.4%) were referred to a pulmonologist, 22 of 76 (28.9%) completed the referral, and 17 of 22 (77.3%) received a new pulmonary diagnosis or change in medications. Of the 155 patients without a history of pulmonary disease, 76 had poor midflow values, 57 (75%) of these patients were referred, and 12 of 57 (21%) completed the referral. Eight (67%) of these 12 patients were diagnosed with a previously unrecognized pulmonary disorder. Of the 44 patients with a prior history of pulmonary disease, 29 (65.9%) had poor midflow values. Nineteen (65.5%) of these patients were referred, and 9 (47%) received a new pulmonary diagnosis or a change in their medications. There were 51 classically trained singers and 148 nonclassically trained singers or nonsingers. There was no significant difference in average midflow values between the two groups (80.96 ± 24.7 and 80.73 ± 28.4, respectively) or in the percentage of classically trained singers with poor midflow values compared with nonsingers (53.5% vs. 49%, respectively).

CONCLUSION

This study suggests that patients with dysphonia may have unrecognized underlying pulmonary disease, and PFT should be considered as part of the routine initial voice evaluation for patients presenting with dysphonia.

Understanding and measuring symptoms and health status in asthma COPD overlap: content validity of the EXACT and SGRQ

BACKGROUND

Asthma-chronic obstructive pulmonary disease overlap (ACO) differs from asthma and chronic obstructive pulmonary disease (COPD) in demographics, phenotypic characteristics and outcomes, yet the patient experience of ACO is poorly characterized. We aimed to understand and compare the patient experience of symptoms and domains of impact in ACO relative to COPD, and assess the content validity of existing patient-reported outcome (PRO) instruments in ACO.

METHODS

This US qualitative, interview study included patients who met American Thoracic Society/European Respiratory Society spirometric criteria for COPD. Additionally, patients with ACO demonstrated reversibility (forced expiratory volume in 1 s [FEV1] increase ≥ 12% and ≥ 200 mL) to albuterol/salbutamol and an FEV1/forced vital capacity ratio < 0.7. Patients took part in concept elicitation (CE) to explore symptoms and impacts of obstructive lung disease. The Exacerbations of Chronic Pulmonary Disease Tool (EXACT), St George's Respiratory Questionnaire (SGRQ) and a daily wheeze assessment were cognitively debriefed to assess relevance and comprehensiveness. Interviews were analyzed using Atlas.Ti. Concept saturation was evaluated at the symptom level.

RESULTS

Twenty patients with ACO and 10 patients with COPD were recruited. Patients from both groups indicated that shortness of breath was their most frequent and bothersome symptom. The most frequently reported symptoms in both groups were shortness of breath, cough, wheezing, difficulty breathing, mucus/phlegm, chest tightness, and tiredness, weakness or fatigue. The onset, severity, frequency and duration of symptoms were consistently described across both groups, although a higher proportion of patients with ACO experienced exacerbations versus those with COPD. Impacts on daily living, physical impacts and emotional impacts were commonly described (ACO: 90-100%, COPD: 80-100%). Concept saturation was achieved in both groups. Overall, the EXACT, SGRQ and daily wheeze assessment were well understood and relevant to most patients with ACO or COPD (50-100%) and patients generally found the assessments easy to complete. The PRO instruments adequately captured symptoms described during CE, demonstrating high content validity in ACO and COPD.

CONCLUSIONS

Patients with ACO and COPD experienced similar symptoms and impacts. The EXACT, SGRQ and assessment of wheeze were well understood and captured concepts relevant to patients with ACO.

Voice disorder in systemic lupus erythematosus

Systemic lupus erythematosus (SLE) is a chronic disease characterized by progressive tissue damage. In recent decades, novel treatments have greatly extended the life span of SLE patients. This creates a high demand for identifying the overarching symptoms associated with SLE and developing therapies that improve their life quality under chronic care. We hypothesized that SLE patients would present dysphonic symptoms. Given that voice disorders can reduce life quality, identifying a potential SLE-related dysphonia could be relevant for the appraisal and management of this disease. We measured objective vocal parameters and perceived vocal quality with the GRBAS (Grade, Roughness, Breathiness, Asthenia, Strain) scale in SLE patients and compared them to matched healthy controls. SLE patients also filled a questionnaire reporting perceived vocal deficits. SLE patients had significantly lower vocal intensity and harmonics to noise ratio, as well as increased jitter and shimmer. All subjective parameters of the GRBAS scale were significantly abnormal in SLE patients. Additionally, the vast majority of SLE patients (29/36) reported at least one perceived vocal deficit, with the most prevalent deficits being vocal fatigue (19/36) and hoarseness (17/36). Self-reported voice deficits were highly correlated with altered GRBAS scores. Additionally, tissue damage scores in different organ systems correlated with dysphonic symptoms, suggesting that some features of SLE-related dysphonia are due to tissue damage. Our results show that a large fraction of SLE patients suffers from perceivable dysphonia and may benefit from voice therapy in order to improve quality of life.