Text-to-speech synthesis as an alternative communication means after total laryngectomy

Use of Speech-to-Text Translation Resources to Address Communication Barriers in Patients With Hearing Loss: A Systematic Review

OBJECTIVE

Patients with concomitant limited English proficiency (LEP) and hearing loss may experience communication barriers, not fully mitigated by traditional interpreter services. Although there is no clear consensus on the most reliable and optimal resources for these patients, speech-to-text (STT) applications with translation capability may bridge these barriers. We review the existing literature applying STT translation programs in clinical settings and identify commercially available STT translation resources to evaluate their prospective application in the otology setting.

DATABASES REVIEWED

PubMed MEDLINE, Embase, and Web of Science.

METHODS

A systematic review of English language peer-reviewed literature was conducted, examining STT translation in clinical settings. An additional search identifying STT software with translation capabilities was completed.

RESULTS

Seven out of 591 unique citations met the inclusion criteria, and 29 unique STT translation applications were identified, supporting up to 140 languages. Stakeholders endorse positive perceptions of STT translation programs. Facilitators to implementation included time and cost feasibility, whereas barriers included risk of mistranslation and inability to assess accuracy. Ongoing machine-learning efforts are underway for developing and improving STT translation technologies, but there is paucity of literature evaluating their application in patients with hearing loss.

CONCLUSIONS

Small studies have suggested an acceptable level of accuracy for STT translation programs as adjunctive to standard of care services. Further work should proactively address implementation challenges with special attention to use of these technologies for patients with concomitant LEP and HL, while supporting additional technological advancement for application of these technologies in otology practice.

Alaryngeal Speech Enhancement for Noisy Environments Using a Pareto Denoising Gated LSTM

Loss of the larynx significantly alters natural voice production, requiring alternative communication modalities and rehabilitation methods to restore speech intelligibility and improve the quality of life of affected individuals. This paper explores advances in alaryngeal speech enhancement to improve signal quality and reduce background noise, focusing on individuals who have undergone laryngectomy. In this study, speech samples were obtained from 23 Lithuanian males who had undergone laryngectomy with secondary implantation of the tracheoesophageal prosthesis (TEP). Pareto-optimized gated long short-term memory was trained on tracheoesophageal speech data to recognize complex temporal connections and contextual information in speech signals. The system was able to distinguish between actual speech and various forms of noise and artifacts, resulting in a 25% drop in the mean signal-to-noise ratio compared to other approaches. According to acoustic analysis, the system significantly decreased the number of unvoiced frames (proportion of voiced frames) from 40% to 10% while maintaining stable proportions of voiced frames (proportion of voiced speech frames) and average voicing evidence (average voice evidence in voiced frames), indicating the accuracy of the approach in selectively attenuating noise and undesired speech artifacts while preserving important speech information.

Implementing a Statistical Parametric Speech Synthesis System for a Patient with Laryngeal Cancer

Total laryngectomy, i.e., the surgical removal of the larynx, has a profound influence on a patient's quality of life. The procedure results in a loss of natural voice, which in effect constitutes a significant socio-psychological problem for the patient. The main aim of the study was to develop a statistical parametric speech synthesis system for a patient with laryngeal cancer, on the basis of the patient's speech samples recorded shortly before the surgery and to check if it was possible to generate speech quality close to that of the original recordings. The recording made use of a representative corpus of the Polish language, consisting of 2150 sentences. The recorded voice proved to indicate dysphonia, which was confirmed by the auditory-perceptual RBH scale (roughness, breathiness, hoarseness) and by acoustical analysis using AVQI (The Acoustic Voice Quality Index). The speech synthesis model was trained using the Merlin repository. Twenty-five experts participated in the MUSHRA listening tests, rating the synthetic voice at 69.4 in terms of the professional voice-over talent recording, on a 0-100 scale, which is a very good result. The authors compared the quality of the synthetic voice to another model of synthetic speech trained with the same corpus, but where a voice-over talent provided the recorded speech samples. The same experts rated the voice at 63.63, which means the patient's synthetic voice with laryngeal cancer obtained a higher score than that of the talent-voice recordings. As such, the method enabled for the creation of a statistical parametric speech synthesizer for patients awaiting total laryngectomy. As a result, the solution would improve the quality of life as well as better mental wellbeing of the patient.

Voice Onset Time in Patients Using Speech Prosthesis After Total Laryngectomy

AIM

Our study was carried out to investigate the presence of known differences in voice and articulation quality after total laryngectomy. Patients provided phonation with tracheoesophageal speech prosthesis. We recorded patients' voice onset time (VOT) values - an important parameter of acoustic analysis.

METHODS

The study included 18 patients with total laryngectomy who received valvular speech prosthesis via a primary or secondary tracheoesophageal fistula between 2009 and 2011 at the Istanbul Training and Research Hospital Otorhinolaryngology Clinic. Twenty healthy male volunteers were included as the control group. All subjects produced the /pa/, /ta/, /ka/ syllables three times, and the VOT values were determined by recording the voices on a computer.

RESULTS

A total of 38 male patients, 18 of which were patients with total laryngectomy and tracheoesophageal speech prosthesis (aged between 46 and 75 years, mean: 59) and 20 controls (aged between 50 and 70 years, mean: 58), were included in the study. The age distribution of the groups did not differ statistically (P > 0.05). In the total laryngectomy and tracheoesophageal speech prosthesis group, the VOT mean values of the /ka/ syllable were significantly lower than the control group, whereas the /pa/ (P = 0.848) and /ta/ VOT mean values (P = 0.809) were similar between groups.

CONCLUSIONS

This study shows that there is no significant difference in the articulation of voiceless plosives, except for the /ka/ sound, between patients using speech prostheses after total laryngectomy and controls. For standardization of these measured values and their use in clinical practice, it may be beneficial to support this study with studies that involve more patients and examine different indicators showing the quality and intelligibility of other voice characteristics.