Search

Background: Patients with limited English proficiency frequently receive substandard health care. Asynchronous telepsychiatry (ATP) has been established as a clinically valid method for psychiatric assessments. The addition of automated speech recognition (ASR) and automated machine translation (AMT) technologies to asynchronous telepsychiatry may be a viable artificial intelligence (AI)-language interpretation option. Objective: This project measures the frequency and accuracy of the translation of figurative language devices (FLDs) and patient word count per minute, in a subset of psychiatric interviews from a larger trial, as an approximation to patient speech complexity and quantity in clinical encounters that require interpretation. Methods: A total of 6 patients were selected from the original trial, where they had undergone 2 assessments, once by an English-speaking psychiatrist through a Spanish-speaking human interpreter and once in Spanish by a trained mental health interviewer-researcher with AI interpretation. 3 (50%) of the 6 selected patients were interviewed via videoconferencing because of the COVID-19 pandemic. Interview transcripts were created by automated speech recognition with manual corrections for transcriptional accuracy and assessment for translational accuracy of FLDs. Results: AI-interpreted interviews were found to have a significant increase in the use of FLDs and patient word count per minute. Both human and AI-interpreted FLDs were frequently translated inaccurately, however FLD translation may be more accurate on videoconferencing. Conclusions: AI interpretation is currently not sufficiently accurate for use in clinical settings. However, this study suggests that alternatives to human interpretation are needed to circumvent modifications to patients' speech. While AI interpretation technologies are being further developed, using videoconferencing for human interpreting may be more accurate than in-person interpreting. [ABSTRACT FROM AUTHOR]

Objective: In the United States, most adults who receive cochlear implants (CIs) do not undergo a comprehensive auditory rehabilitation (CAR) approach, which may result in suboptimal outcomes. The objectives of this pilot study were to demonstrate that a CAR approach incorporating auditory training (AT) by a speech‐language pathologist (SLP) is feasible in adults receiving CIs and to explore whether this approach results in improved outcomes. Methods: Twenty‐four postlingually deaf adult CI candidates were serially assigned to one of three groups: (a) a "CAR group" that received standard of care implantation, programming by an audiologist, an additional preoperative counseling session, and eight one‐hour AT sessions; (b) a "passive control" standard‐of‐care group; and (c) an "active control" group that also received the extra preoperative counseling session. Participants were tested preoperatively and 1, 3, and 6 months after CI using measures of word and sentence recognition in quiet and in babble, as well as measures of quality of life (QOL). Results: The CAR approach was feasible, but this pilot study was underpowered to determine efficacy. Differential time courses of speech recognition improvement were seen for sentence and word recognition. All QOL measurements showed improvement from pre‐CI to 1 month post‐CI activation. Results revealed issues to consider for a larger‐scale study of CAR revolving around participant selection, study measures, and sample size. Conclusion: The CAR approach is feasible in new CI users. A larger trial is needed to investigate whether CAR leads to better outcomes or faster improvement in this clinical population. Level of Evidence: 2. In the United States, most adults who receive cochlear implants (CIs) do not undergo a comprehensive auditory rehabilitation (CAR) approach, which may result in suboptimal outcomes. This pilot study investigated the feasibility and efficacy of applying a CAR approach to new adult CI users. The approach was found to be feasible, and a larger trial is needed to investigate whether CAR leads to better outcomes or faster improvement in this clinical population. [ABSTRACT FROM AUTHOR]

This paper shows that extraction and analysis of various acoustic features from speech using mobile devices can allow the detection of patterns that could be indicative of neurological trauma. This may pave the way for new types of biomarkers and diagnostic tools. Toward this end, we created a mobile application designed to diagnose mild traumatic brain injuries (mTBI) such as concussions. Using this application, data were collected from youth athletes from 47 high schools and colleges in the Midwestern United States. In this paper, we focus on the design of a methodology to collect speech data, the extraction of various temporal and frequency metrics from that data, and the statistical analysis of these metrics to find patterns that are indicative of a concussion. Our results suggest a strong correlation between certain temporal and frequency features and the likelihood of a concussion. [ABSTRACT FROM AUTHOR]

Background: Natural variability in speech is a significant challenge to robust successful spoken word recognition. In everyday listening environments, listeners must quickly adapt and adjust to multiple sources of variability in both the signal and listening environments. High-variability speech may be particularly difficult to understand for non-native listeners, who have less experience with the second language (L2) phonological system and less detailed knowledge of sociolinguistic variation of the L2. Purpose: The purpose of this study was to investigate the effects of high-variability sentences on nonnative speech recognition and to explore the underlying sources of individual differences in speech recognition abilities of non-native listeners. Research Design: Participants completed two sentence recognition tasks involving high-variability and low-variability sentences. They also completed a battery of behavioral tasks and self-report questionnaires designed to assess their indexical processing skills, vocabulary knowledge, and several core neurocognitive abilities. Study Sample: Native speakers of Mandarin (n = 25) living in the United States recruited from the Indiana University community participated in the current study. A native comparison group consisted of scores obtained from native speakers of English (n = 21) in the Indiana University community taken from an earlier study. Data Collection and Analysis: Speech recognition in high-variability listening conditions was assessed with a sentence recognition task using sentences from PRESTO (Perceptually Robust English Sentence Test Open-Set) mixed in 6-talker multitalker babble. Speech recognition in low-variability listening conditions was assessed using sentences from HINT (Hearing In Noise Test) mixed in 6-talker multitalker babble. Indexical processing skills were measured using a talker discrimination task, a gender discrimination task, and a forced-choice regional dialect categorization task. Vocabulary knowledge was assessed with the WordFam word familiarity test, and executive functioning was assessed with the BRIEF-A (Behavioral Rating Inventory of Executive Function - Adult Version) self-report questionnaire. Scores from the non-native listeners on behavioral tasks and self-report questionnaires were compared with scores obtained from native listeners tested in a previous study and were examined for individual differences. Results: Non-native keyword recognition scores were significantly lower on PRESTO sentences than on HINT sentences. Non-native listeners’ keyword recognition scores were also lower than native listeners’ scores on both sentence recognition tasks. Differences in performance on the sentence recognition tasks between non-native and native listeners were larger on PRESTO than on HINT, although group differences varied by signal-to-noise ratio. The non-native and native groups also differed in the ability to categorize talkers by region of origin and in vocabulary knowledge. Individual non-native word recognition accuracy on PRESTO sentences in multitalker babble at more favorable signal-to-noise ratios was found to be related to several BRIEF-A subscales and composite scores. However, non-native performance on PRESTO was not related to regional dialect categorization, talker and gender discrimination, or vocabulary knowledge. Conclusions: High-variability sentences in multitalker babble were particularly challenging for nonnative listeners. Difficulty under high-variability testing conditions was related to lack of experience with the L2, especially L2 sociolinguistic information, compared with native listeners. Individual differences among the non-native listeners were related to weaknesses in core neurocognitive abilities affecting behavioral control in everyday life. [ABSTRACT FROM AUTHOR]

Spoken language recognition refers to the automatic process through which we determine or verify the identity of the language spoken in a speech sample. We study a computational framework that allows such a decision to be made in a quantitative manner. In recent decades, we have made tremendous progress in spoken language recognition, which benefited from technological breakthroughs in related areas, such as signal processing, pattern recognition, cognitive science, and machine learning. In this paper, we attempt to provide an introductory tutorial on the fundamentals of the theory and the state-of-the-art solutions, from both phonological and computational aspects. We also give a comprehensive review of current trends and future research directions using the language recognition evaluation (LRE) formulated by the National Institute of Standards and Technology (NIST) as the case studies. [ABSTRACT FROM PUBLISHER]

Abstract: This study attempts to identify strategies peculiar to speakers of non-standard varieties of American English when composing with speech recognition-based dictation software. Two speakers of African American Vernacular English and one non-native speaker of English were asked to orally compose a series of college-level writing tasks using ViaVoice 98. The findings indicate that when the software misrecognized their speech, the language minority students resorted to similar strategies as those of standard American speakers discussed in the literature such as speaker adaptation to recognition systems leading to hyperarticulation and other types of phonetic and lexical adjustments. Moreover, the participants based their strategies on their knowledge of human-to-human interaction (e.g., the sequential organization of talk), which more often than not proved to be useless when interacting with the software. Yet, the students did seem to build up an awareness during the speech sessions of their linguistic variety and its divergence from the standard American accent that the software was originally modeled after. The findings of this study offer a better understanding of some of the strategies that speakers from diverse ethno-linguistic backgrounds may employ when speaking to write with speech recognition-based dictation software. They also offer some insight into users’ assumptions about the capabilities of the system, i.e., what they believed the speech-based software could process in terms of utterance length, tone of voice, canonical pronunciation, and lexical items. [Copyright &y& Elsevier]

Cochlear implant recipients have demonstrated remarkable increases in speech perception since US FDA approval was granted in 1984. Improved performance is due to a number of factors including improved cochlear implant technology, evolving speech coding strategies, and individuals with increasingly more residual hearing receiving implants. Despite this evolution, the same recommendations for pre- and postimplant speech recognition testing have been in place for over 10 years in the United States. To determine whether new recommendations are warranted, speech perception performance was assessed for 156 adult, postlingually deafened implant recipients as well as 50 hearing aid users on monosyllabic word recognition (CNC) and sentence recognition in quiet (HINT and AzBio sentences) and in noise (BKB-SIN). Results demonstrated that for HINT sentences in quiet, 28% of the subjects tested achieved maximum performance of 100% correct and that scores did not agree well with monosyllables (CNC) or sentence recognition in noise (BKB-SIN). For a more difficult sentence recognition material (AzBio), only 0.7% of the subjects achieved 100% performance and scores were in much better agreement with monosyllables and sentence recognition in noise. These results suggest that more difficult materials are needed to assess speech perception performance of postimplant patients – and perhaps also for determining implant candidacy. Copyright © 2008 S. Karger AG, Basel [ABSTRACT FROM AUTHOR]

Larson et al (2000) reported the findings of a multicenter, NIDCD/VA clinical trial that compared hearing aid performance for three output limiting circuits in 360 adults with symmetrical sensorineural hearing loss. The current study was undertaken to examine long-term hearing aid benefit in this same group of participants following five to six years of hearing aid use. The speech-recognition portion of the follow-up study enrolled 108 participants from the original study, 85% of whom were current hearing aid users and 15% of whom had not worn hearing aids during the past month (nonusers). Recognition performance in sound field on the NU-6 (quiet at 62 dB SPL) and the CST (quiet at 74 dB SPL and with -3 and 3 dB signal-to-babble ratios [S/B] at 62 and 74 dB SPL) was measured unaided and aided whenever possible. Speech-recognition abilities decreased significantly since the original study. Speech-recognition decrements were observed regardless of the speech materials (NU-6 and CST), test condition (quiet and noise), S/B (-3 and 3 dB), or stimulus level (62 and 74 dB SPL). Despite decreases in speech recognition, hearing aid benefit remained largely unchanged since the original study; aided performance exceeded unaided performance regardless of presentation level or noise condition. As in the original study, the relations among stimulus level, S/B, and speech-recognition performance were complex. [ABSTRACT FROM AUTHOR]

The article reports on the results of an exploratory study concerning the training time, accuracy and end-user acceptance of Speech Recognition Systems (SRS) technology in the U.S. SRS offers many possibilities with its ability to understand voice input. In addition, it will convert voice input into text that the computer can understand.

This work presents research analyzing the structure of how U.S. addresses are spoken by Spanish speakers living in the U.S. The findings, based on live-call recordings of Spanish speakers interacting with call center agents, reveal that an English-like format with the street number spoken before the street name was used in 91% of the cases, even though the street number was said in Spanish 100% of the time. How audio prompt wording can be used to influence the way addresses are spoken and also how street names consisting of numbers are said are possible areas for future study. In addition, some best practices related to dialog design for bilingual Spanish-English speech recognition applications are also revealed during this study. [ABSTRACT FROM AUTHOR]

With the increasing prevalence of PACS over the past decade, face-to-face image review among health care providers has become a rarity. This change has resulted in increasing dependence on fast and accurate communication in radiology. Turnaround time expectations are now conveyed in minutes rather than hours or even days. Ideal modern radiology communication is a closed-loop cycle with multiple interoperable applications contributing to the final product. The cycle starts with physician order entry, now often performed through the electronic medical record, with clinical decision support to ensure that the most effective imaging study is ordered. Radiology reports are now almost all in electronic format. The majority are produced using speech recognition systems. Optimization of this software use can alleviate some, if not all, of the inherent user inefficiencies in this type of reporting. Integrated third-party software applications that provide data mining capability are extremely helpful in both academic and clinical settings. The closed-loop ends with automated communication of imaging results. Software products for this purpose should facilitate use of levels of alert, automated escalation to providers, and recording of audit trails of reports received. The multiple components of reporting should be completely interoperable with each other, as well as with the PACS, the RIS, and the electronic medical record. This integration will maximize radiologist efficiency and minimize the possibility of communication error., (Copyright © 2014. Published by Elsevier Inc.)