A Deep Learning Approach for Voice Disorder Detection for Smart Connected Living Environments

Edge Analytics and Artificial Intelligence are important features of the current smart connected living community. In a society where people, homes, cities, and workplaces are simultaneously connected through various devices, primarily through mobile devices, a considerable amount of data is exchanged, and the processing and storage of these data are laborious and difficult tasks. Edge Analytics allows the collection and analysis of such data on mobile devices, such as smartphones and tablets, without involving any cloud-centred architecture that cannot guarantee real-time responsiveness. Meanwhile, Artificial Intelligence techniques can constitute a valid instrument to process data, limiting the computation time, and optimising decisional processes and predictions in several sectors, such as healthcare. Within this field, in this article, an approach able to evaluate the voice quality condition is proposed. A fully automatic algorithm, based on Deep Learning, classifies a voice as healthy or pathological by analysing spectrogram images extracted by means of the recording of vowel /a/, in compliance with the traditional medical protocol. A light Convolutional Neural Network is embedded in a mobile health application in order to provide an instrument capable of assessing voice disorders in a fast, easy, and portable way. Thus, a straightforward mobile device becomes a screening tool useful for the early diagnosis, monitoring, and treatment of voice disorders. The proposed approach has been tested on a broad set of voice samples, not limited to the most common voice diseases but including all the pathologies present in three different databases achieving F1-scores, over the testing set, equal to 80%, 90%, and 73%. Although the proposed network consists of a reduced number of layers, the results are very competitive compared to those of other “cutting edge” approaches constructed using more complex neural networks, and compared to the classic deep neural networks, for example, VGG-16 and ResNet-50.

Usefulness of Vocal Fatigue Index for Hypertension of Extrinsic Laryngeal Muscles

Background and Objectives This study compares Vocal Fatigue Index (VFI) scores according to the presence or absence of external laryngeal tension in hyperfunctional voice disorder. And through this, it is to confirm the usefulness of VFI to hypertension of extrinsic laryngeal muscles.Materials and Method The subjects were 61 female diagnosed with hyperfunctional voice disorder (hypertension group 41, non-hypertension group 20). The author palpated extrinsic laryngeal muscles for evaluation of hypertension and classified them as the presence or absence. The voice measurements were jitter, shimmer, Korean-Voice Handicap Index-10 (K-VHI-10), and Korean-Vocal Fatigue Index (K-VFI). The voice compared were according to the diagnosis and presence of hypertension only for patients with hyperfunctional voice disorder.Results As a result of comparing the voice measurement according to the presence or absence of hypertension, there was no significant difference in the acoustic variables, K-VHI-10 and K-VFI-Total, K-VFI-Fatigue. Whereas, K-VFI-Physical (p=0.006) and K-VFI-Rest (p=0.022) were significantly higher in the hypertension group.Conclusion These results indicate that the hypertension group has more physical discomfort and less voice recovery than the group without hypertension. It means that K-VFI can measure the physical discomfort and limitations of voice recovery due to hypertension of the external laryngeal muscle. The VFI can be used as one of the methods to evaluate the hypertension of the external laryngeal muscle in Hyperfunctional voice disorder.

Botulinum Toxin Therapy: A Series of Clinical Studies on Patients with Spasmodic Dysphonia in Japan

Spasmodic dysphonia (SD) is a rare voice disorder caused by involuntary and intermittent spasms of the laryngeal muscles. Both diagnosis and treatment have been controversial. Therefore, a series of clinical studies has recently been conducted in Japan. A nationwide epidemiological survey revealed that adductor SD predominated (90–95% of all cases; 3.5–7.0/100,000), principally among young women in their 20s and 30s. To facilitate early diagnosis, we created diagnostic criteria for SD and a severity grading system. The diagnostic criteria include the principal and accompanying symptoms, clinical findings during phonation, the treatment response, and the differential diagnoses. The severity grade is determined using a combination of subjective and objective assessments. Botulinum toxin (BT) injection is the treatment of choice; however, there have been few high-quality clinical studies and BT has been used off-label. We conducted a placebo-controlled, randomized, double-blinded clinical trial of BT therapy; this was effective and safe. BT treatment is now funded by the Japanese medical insurance scheme. Studies thus far have facilitated early diagnosis and appropriate therapy; they have fostered patient awareness of SD.

Relationships Across Clinical Measures of Vocal Quality and Functioning and Their Relationship With Patient Perception

Purpose The purpose of this study was to investigate the relationships among subjective auditory-perceptual ratings of vocal quality, objective acoustic and aerodynamic measures of vocal function, and patient-perceived severity of their vocal complaint. Method This study was a retrospective chart review of adult patients evaluated at a single outpatient center over a 1.5-year time period. Twenty-two clinical objective and subjective measures of voice were extracted from 676 charts (310 males, 366 females). To identify the underlying concepts addressed in an initial voice assessment, principal component analyses were conducted for males and females to account for sex differences. Linear regression models were conducted to examine the relationship between the principal components and patient perceived severity. Results Seven principal components were identified for both sexes and accounted for 75% and 71% of the variance in the clinical measures, respectively. Of these seven principal components, only two predicted male patient perceived severity, which accounted for 22% of the variance. In contrast, four principal components predicted female patient perceived severity of their voice disorder and accounted for 19% of the variance. Conclusions The results highlight the underlying aspects of vocal quality and functioning that are evaluated during an initial assessment. Male and female patients differ in which of these components may contribute self-perceived severity of a voice disorder. Identifying these underlying components may support clinical decision making when developing a clinical protocol and highlights the overlap between patient concerns and clinical measures. Supplemental Material https://doi.org/10.23641/asha.16879603

EP.WE.852Role of Prophylactic steroids in the prevention of post-thyroidectomy hypocalcaemia and voice dysfunction

Aim Total thyroidectomy, being a common surgical option, for multinodular goitre and selective cases of Graves’ disease, is associated with risks related to temporary hypocalcaemia and vocal quality dysfunction. Dexamethasone, as a steroid, has well-established anti-inflammatory and immune-modulating effects. As researchers continue to understand its pathophysiology on hypocalcaemia and voice quality, we aim to assess the effect of preoperative dexamethasone used to improve hypocalcaemia and vocal cord function/post-thyroidectomy voice disorder. Methods We conducted a single-centre, non-randomised, double-blinded placebo-controlled, parallel-group study with short follow-up in a public sector hospital from January 2014 to December 2019. Patients were equally distributed in two groups to compare the effect of dexamethasone with placebo on hypocalcaemia and vocal cord function/post-thyroidectomy voice disorder Results Out of 192 patients, 36 (18.75) male patients while 156 (81.25) were female. Age range was 18 to 60. Eighteen patients had both low calcium level at 24 hours after the operation and hypocalcemic symptoms, of which 4 remain to have hypocalcaemia on third postoperative day. Forty patients defined as having voice dysfunction at 24 hours after being scored using Voice scale. Of these 40 patients with voice dysfunction, 8 had low calcium level in blood at 24 hours while only 2 of these remained with low calcium level on third postoperative day. 6/8 patients had symptomatic hypocalcaemia on first postoperative day. Conclusions Single-dose dexamethasone, during the preoperative period, is as safe and as effective in improving transient postoperative hypocalcaemia and Temporary vocal quality dysfunction/post-thyroidectomy voice disorder.

Characterization of Cervical-Cranial Muscle Network in Correlation with Vocal Features

Objective evaluation of physiological responses using non-invasive methods has attracted great interest regarding the assessment of vocal performance and disorders. This paper, for the first time, demonstrates that the topographical features of the cervical-cranial intermuscular coherence network generated using surface electromyography (sEMG) have a strong potential for detecting subtle changes in vocal performance. For this purpose, in this paper, 12 sEMG signals were collected from six cervical and cranial muscles bilaterally. Data were collected from four subjects without a history of a voice disorder performing a series of vocal tasks. The vocal tasks were varied phonation (an /a/ sustained for the maximal duration with combinations of two levels of loudness and two levels of pitch), a pitch glide from low to high, singing a familiar song, spontaneous speech, and reading with different loudness levels. The varied phonation tasks showed the median degree, and weighted clustering coefficient of the coherence-based intermuscular network ascends monotonically, with a high effect size (|rrb| = 0.52). The set of tasks, including pitch glide, singing, and speech, was significantly distinguishable using the network features as both degree and weighted clustering coefficient had a very high effect size (|rrb| > 0.83) across these tasks. Also, pitch glide has the highest degree and weighted clustering coefficient among all tasks (degree > 0.6, weighted clustering coefficient > 0.6). Spectrotemporal features performed far less effective than the proposed functional muscle network metrics to differentiate the vocal tasks. The highest effect size for spectrotemporal features was only |rrb| = 0.19. In this paper, for the first time, the power of a cervical-cranial muscle network has been demonstrated as a neurophysiological window to vocal performance. The results also shed light on the tasks with the highest network involvement, which may be potentially used in monitoring vocal disorders and tracking rehabilitation progress.

Intractable Muscle Tension Dysphonia Treated by Injection Laryngoplasty and Lidocaine Injection

Muscle tension dysphonia (MTD) is a voice disorder characterized by excessive tension of the laryngeal muscles during phonation. Voice therapy is the gold standard of treatment for MTD. However, patients with MTD do not always respond to voice therapy. Multidisciplinary approaches have been attempted to treat intractable MTD such as lidocaine instillation, lidocaine injection to recurrent laryngeal nerve, botox injection and excision of false ventricle using CO2 laser. Recently, injection laryngoplasty is suggested that assists in more efficient phonation and voice therapy to MTD patients. A patient with intractable MTD underwent lidocaine injection and injection laryngoplasty showed improved voice quality and remained stable until postoperative 3 months without any complications.