Search

GLUT1 deficiency syndrome (GLUT1DS) is a rare genetic disorder caused by a mutation in the SLC2A1 gene that limits the transport of glucose across the blood-brain barrier. Speech disorders and dysarthria are typical findings in patients with GLUT1DS, but have never been deeply phenotyped. The aim of the present study was to characterize speech abilities in a sample of patients with GLUT1DS. 30 patients with GLUT1DS were recruited. We reported impairments in different speech and oromotor domains: the speech was characterized by dysarthria, inaccurate articulation of consonants, abnormal nasal resonance, errors in intonation and prosody and low intelligibility. We observed difficulties in motor planning and programming. Moreover, we observed a significant difference between the dysarthric level of impairment with genotype groups. The presence of a speech disorder in patients with GLUT1DS represents a core feature of the syndrome. Our findings suggest that patients with GLUT1DS would benefit from a comprehensive neurocognitive assessment to detect strengths and weaknesses of the speech profile. Understanding the speech and language phenotype in GLUT1DS is critical for planning early intervention to positively influence the global development of patients with GLUT1DS. • The purpose of this study is to analyze and characterize the speech profile of a sample of patients with GLUT1DS. • The majority of the patients manifested dysarthria with different levels of severity. • A relationship between genotype and dysartric level of impairment was found. • The introduction of the ketogenic diet therapy might be a protective factor for cognitive outcomes in patients with GLUT1DS. [ABSTRACT FROM AUTHOR]

The recent research trend advocates the application of new technological advances in therapy and rehabilitation. Applications of modern technologies are good candidates because they have the potential to prevent many diseases and engage young children in therapeutic interventions. Regardless of previous findings, novel approaches and planned strategic efforts can help several people with disabilities achieve a normal lifestyle by using 2D and 3D games in therapeutic intervention. As we know, the design of biofeedback-based therapeutic games is full of challenges, from data collection to standardization and commercialization. Since this field requires periodic development from Time to Time due to rapidly changing technology and societal demands, this article explains the background of the therapeutic benefits associated with play therapies and their effectiveness in children. This article examines alterations in children's performance by modifying various aspects of gameplay. Throughout the study, both a blinded physiotherapist and a training physiotherapist were involved. Parameters such as biting, spoon-feeding, chewing, drinking, and swallowing were adjusted and observed. The Functional Feeding Assessment (FFA) and the Bayley Scales of Infant Development (BSID-II) scores were then analyzed and compared between groups based on their respective scoring systems. By comparing the training group and control group for the given parameters, we got p-values of Biting 0.027, Spoon Feeding 0.047, Chewing 0.040, Drinking 0.050, Swallowing 0.060, and t- value Biting 0.092, Spoon Feeding 0.463, Chewing 0.856, Drinking 0.118, Swallowing 0.131. Children's behavior to play on computers and mobile is transformed into valuable applications for preventing, treating, and training children with oral motor and speech disorders. This gaming therapy benefits children with oral motor skills and CP disability. [ABSTRACT FROM AUTHOR]

Speech disorders encompass a diverse range of conditions hindering effective communication, stemming from developmental, neurological, or physical factors. These challenges impact daily life and opportunities. Automatic Speech Recognition (ASR) technology emerges as a pivotal solution. ASR, powered by sophisticated algorithms, transcribes spoken language into written text, transcending traditional communication barriers. Tailorable to accommodate unique speech disorders, ASR offers the promise of empowering individuals to articulate their thoughts. Our Python Flask-based ASR(FBASR) applications meticulously crafted to cater to individuals, including children, grappling with speech disorders. It provides a user-friendly platform. Users simply submit a speech audio file in WAV format, and the application transcribes it. Alongside recognized text, it furnishes vital audio metrics. Additionally, it generates visual audio signal and spectrogram representations, all stored efficiently for future reference. The application's profound advantage lies in granting a voice to individuals with speech disorders, facilitating education, employment, and healthcare interactions. Furthermore, it serves as a resource for research and therapy development, promising inclusivity and enriching lives. The proposed work concludes by discussing the future outlook of cloud-based ASR and identifying some of the key areas of research that need to be addressed to make cloud-based DSR systems more accurate, reliable, and accessible. [ABSTRACT FROM AUTHOR]

This review provides a comprehensive overview of the literature on velopharyngeal insufficiency, associated anomalies, and speech/language impairment in patients with craniofacial microsomia (CFM). A systematic search of the literature was conducted to identify records on VPI and speech impairment in CFM from their inception until September 2022 within the databases Embase, PubMed, MEDLINE, Ovid, CINAHL EBSCO, Web of Science, Cochrane, and Google Scholar. Seventeen articles were included, analysing 1,253 patients. Velopharyngeal insufficiency results in hypernasality can lead to speech impairment. The reported prevalence of both velopharyngeal insufficiency and hypernasality ranged between 12.5% and 55%, while the reported prevalence of speech impairment in patients with CFM varied between 35.4% and 74%. Language problems were reported in 37% to 50% of patients. Speech therapy was documented in 45.5% to 59.6% of patients, while surgical treatment for velopharyngeal insufficiency consisted of pharyngeal flap surgery or pharyngoplasty and was reported in 31.6% to 100%. Cleft lip and/or palate was reported in 10% to 100% of patients with CFM; these patients were found to have worse speech results than those without cleft lip and/or palate. No consensus was found on patient characteristics associated with an increased risk of velopharyngeal insufficiency and speech/language impairment. Although velopharyngeal insufficiency is a less commonly reported characteristic of CFM than other malformations, it can cause speech impairment, which may contribute to delayed language development in patients with CFM. Therefore, timely recognition and treatment of speech impairment is essential. [ABSTRACT FROM AUTHOR]

The treatment of head and neck cancers can have a significant impact on function and appearance, which results in impairment of quality of life. Various long-term sequelae of treatment include speech and swallowing difficulty, oral incompetence, trismus, xerostomia, dental caries and osteoradionecrosis. Management has evolved from single modality treatment of either surgery or radiation to multimodality treatment to attain acceptable functional outcomes. Brachytherapy, also known as interventional radiotherapy, with its ability to deliver high doses centrally to the target, has been shown to improve local control rates. Due to the rapid fall-off of the dose of brachytherapy there is better organ at risk sparing as compared with that of external beam radiotherapy. In the head and neck region, brachytherapy has been practiced in various sites, including the oral cavity, oropharynx, nasopharynx, nasal vestibule and paranasal sinuses. Additionally, brachytherapy has also been considered for reirradiation as a salvage. Brachytherapy can also be considered as a perioperative technique together with surgery. Close multidisciplinary cooperation is required for a successful brachytherapy programme. In oral cavity cancers, depending on the location of the tumour, brachytherapy has been shown to preserve oral competence, maintain tongue mobility, maintain speech, swallowing and the hard palate. Brachytherapy in oropharyngeal cancers has been shown to reduce xerostomia and also reduce dysphagia and aspiration post-radiation therapy. For the nasopharynx, paranasal sinus and nasal vestibule, brachytherapy preserves the respiratory function of the mucosa. Despite such an incomparable impact on function and organ preservation, brachytherapy is an underutilised technique for head and neck cancers. There is a strong need to improve brachytherapy utilisation in head and neck cancers. • Brachytherapy results in function preservation in head and neck cancer. • Function preservation has been achieved in various subsites, such as oral cavity, oropharynx, nasopharynx, paranasal sinuses and nasal vestibule. • Brachytherapy with or without external beam irradiation should be integrated in the routine management of head and neck cancers for optimal outcomes. [ABSTRACT FROM AUTHOR]

• Left-hemisphere stroke is associated with impaired vocal sensorimotor mechanisms. • Aberrant neural oscillations predict vocal sensorimotor deficit in stroke survivors. • Normalization oscillatory brain activities improve vocal communication in stroke. The present study investigated the neural oscillatory correlates of impaired vocal sensorimotor control in left-hemisphere stroke. Electroencephalography (EEG) signals were recorded from 34 stroke and 46 control subjects during speech vowel vocalization and listening tasks under normal and pitch-shifted auditory feedback. Time-frequency analyses revealed aberrantly decreased theta (4–8 Hz) and increased gamma band (30–80 Hz) power in frontal and posterior parieto-occipital regions as well as reduced alpha (8–13 Hz) and beta (13–30 Hz) desynchronization over sensorimotor areas before speech vowel vocalization in left-hemisphere stroke compared with controls. Subjects with the stroke also presented with aberrant modulation of broadband (4–80 Hz) neural oscillations over sensorimotor regions after speech vowel onset during vocalization and listening under normal and altered auditory feedback. We found that the atypical pattern of broadband neural oscillatory modulation was correlated with diminished vocal feedback error compensation behavior and the severity of co-existing language-related aphasia symptoms associated with left-hemisphere stroke. These findings indicate complex interplays between the underlying mechanisms of speech and language and their deficits in post-stroke aphasia. Our data motivate the notion of studying neural oscillatory dynamics as a critical component for the examination of speech and language disorders in post-stroke aphasia. [ABSTRACT FROM AUTHOR]

There remains a paucity of evidence with regards to functional outcomes following the reconstruction of segmental defects in the mandible. It is, however, well recognised that oral rehabilitation following head and neck surgery is a driver of improved quality of life outcomes. We present a prospective service review of functional outcomes of a consecutive cohort of patients following segmental mandibulectomy and virtual surgical planning (VSP) composite fibular free-flap reconstruction. Twenty-five patients, who were identified as having a complete dataset with a minimum of 12 months' follow up, ultimately met the inclusion criteria. Validated functional outcome measures were used primarily to assess speech, diet, and swallowing outcomes. The results demonstrate a decline in both speech and swallowing outcomes at three months postoperatively, with a decline of 37% in the Speech Handicap Index from the preoperative baseline, and a decline of 35% in the MD Anderson Dysphagia Inventory score over the same period. The MD Anderson Dysphagia Inventory score improved at 12 months, whereas the Speech Handicap Index did not. Fundamentally a collaborative approach is required between members of the multidisciplinary team (MDT) to enable optimal patient outcomes. [ABSTRACT FROM AUTHOR]

Stuttering is a speech disorder during which the flow of speech is interrupted by involuntary pauses and repetition of sounds. Stuttering identification is an interesting interdisciplinary domain research problem involving pathology, psychology, acoustics, and signal processing, making it hard and complicated to detect. Recent developments in the machine and deep learning have dramatically revolutionized the speech domain, however minimal attention has been given to stuttering identification. This work fills the gap by trying to bring researchers together from interdisciplinary fields. In this paper, we comprehensively review acoustic features, and statistical and deep learning-based stuttering/disfluency classification methods. We also present several challenges and possible future directions. [ABSTRACT FROM AUTHOR]

For patients with a dentofacial deformity undergoing a planned segmentation of the maxilla for the management of a transverse maxillary arch discrepancy, palatal mucosa tearing may occur during sawing or palatal expansion traction, giving rise to an oronasal communication. This technical note describes the covering of a tear in the palatal mucosa using a buccal fat pad (BFP) flap, in the context of maxillary segmentation during Le Fort I osteotomy. Through the limited buccal incision used for the Le Fort I osteotomy, a small incision is made in the right periosteum posteriorly, and a supraperiosteal dissection is performed to access the BFP. After a sufficient amount of flap is made available, it is gently introduced through the osteotomy gap until it reaches the palatal mucosa defect and is then sutured. In the patient case presented, the palatal mucosa healed fully within 18 days, and the patient reported no nasal regurgitation of food, defective speech, fetid odour, bad taste, or upper respiratory tract or ear infection during the postoperative period. This technique using a BFP flap should therefore be considered in the context of unexpected tearing of the palatal mucosa in patients undergoing a segmented Le Fort I osteotomy. [ABSTRACT FROM AUTHOR]

The Public Opinion Survey of Human Attributes - Stuttering (POSHA-S, St. Louis, 2013) was developed as a standard measure of public attitudes about people who stutter. As with any survey-based methods, threats to validity may occur because of social desirability bias. Using computer mouse-tracking, we were interested in observing changes in cognition that are manifested in intentionality through action by evaluating underlying cognitive processes that drive social judgments of people who stutter. Twenty-two women, 1 non-binary person, and 47 men reported using a computer mouse to complete an online, remote, and modified version of the POSHA-S. Responses were categorized as correct/helpful or incorrect/unhelpful relative to each component of the POSHA-S and were used as measures of explicit cognitive processes. Computer-mouse trajectory metrics, including area under the curve (AUC) and reaction time (RT), were used to measure implicit cognitive processes. Although participants' explicit responses were significantly more likely to be correct/helpful than incorrect/unhelpful, with endorsement of correct/helpful prompts 77 % of the time, participants also endorsed incorrect/unhelpful prompts more than half (i.e., 52 %) of the time. Familiarity with people who stutter was associated with disagreeing with incorrect/unhelpful prompts. As indicated by greater AUC, participants exhibited significantly more implicit cognitive processes indicating competition when responding "disagree" compared to "agree", regardless of whether the prompts were correct/helpful or incorrect/unhelpful. Similarly, participants took significantly longer to respond to prompts with "disagree" rather than "agree". The findings of this study offer evidence of participants reporting cognitive processes that are overall more correct/helpful than incorrect/unhelpful, in their explicit responses to the dichotomous response tasks of the POSHA-S. However, these findings are tempered by evidence of a tendency to agree with statements in the measure and suggest the need for further research to increase understanding of how to measure and improve explicit and implicit cognitive processes related to people who stutter. • Mouse-tracking was used to measure stuttering-related explicit and implicit cognitive processes. • Participants endorsed helpful/correct prompts 77 % of the time. • Participants endorsed unhelpful/incorrect prompts 52 % of the time. • Greater uncertainty was apparent when participants disagreed with statements. [ABSTRACT FROM AUTHOR]

The present study investigated how instructions for paying attention to auditory feedback may affect speech error detection and sensorimotor control. Electroencephalography (EEG) and speech signals were recorded from 21 neurologically intact adult subjects while they produced the speech vowel sound /a/ and received randomized ±100 cents pitch-shift alterations in their real-time auditory feedback. Subjects were instructed to pay attention to their auditory feedback and press a button to indicate whether they detected a pitch-shift stimulus during trials. Data for this group was compared with 22 matched subjects who completed the same speech task under altered auditory feedback condition without attentional instructions. Results revealed a significantly smaller magnitude of speech compensations in the attentional-instruction vs. no-instruction group and a positive linear association between the magnitude of compensations and P2 event-related potential (ERP) amplitudes. In addition, we found that the amplitude of P2 ERP component was significantly larger in the attentional-instruction vs. no-instruction group. Source localization analysis showed that this effect was accounted for by significantly stronger neural activities in the right hemisphere insula, precentral gyrus, postcentral gyrus, transverse temporal gyrus, and superior temporal gyrus in the attentional-instruction group. These findings suggest that attentional instructions may enhance speech auditory feedback error detection, and subsequently improve sensorimotor control via generating more stable speech outputs (i.e., smaller compensations) in response to pitch-shift alterations. Our data are informative for advancing theoretical models and motivating targeted interventions with a focus on the role of attentional instructions for improving treatment outcomes in patients with motor speech disorders. • Attentional instructions improve speech auditory feedback error correction mechanisms. • Attentional instructions reduce speech compensation to auditory feedback alterations. • Neural responses to speech feedback error are enhanced following attentional instructions. • Long-latency neural responses correlate with the magnitude of speech error compensation. • Attention instructions modulate neural activity in right hemisphere sensorimotor networks. [ABSTRACT FROM AUTHOR]

Emerging research shows that children and young adults (CYAs) with cerebral palsy (CP) are at higher risk for attention-deficit/hyperactivity disorder (ADHD). However, little is known about the clinical and functional characteristics of CYAs with these co-occurring disorders. To estimate associations between a diagnosis of ADHD among CYAs with CP and clinical and functional characteristics. This retrospective, cross-sectional study used data from the electronic health records of CYAs (aged 4–26 years) with CP (n = 1145). We used bivariate and multivariable analyses to estimate associations between an ADHD diagnosis, CP type, Gross Motor Function Classification System (GMFCS) level, speech or language disorder, and intellectual disability. 18.1 % of CYAs with CP had a diagnosis of ADHD. CYAs with spastic-bilateral CP had lower odds of ADHD (adjusted odds ratio [AOR] = 0.58; 95 % confidence interval [CI], 0.35–0.96). Odds of having ADHD were significantly lower for those with GMFCS levels III-V (AOR = 0.10; 95 % CI, 0.06–0.15). Our study found that a diagnosis of ADHD among CYAs with CP was associated with greater clinical and functional impairments compared to counterparts without ADHD. Findings highlight the need to screen for both conditions because of the high comorbidity rates in this population. • Prevalence of ADHD is higher in children and young adults (CYAs) with CP. • CYAs with spastic-bilateral CP had lower odds of an ADHD diagnosis. • Odds of ADHD were lower for those with severe physical limitations. [ABSTRACT FROM AUTHOR]

• Understanding why stuttering disappears or is markedly reduced in some special conditions can help understand why it occurs in normal conditions. • Fluency-enhancing conditions draw the speaker's attention to the auditory feedback of speech. • Increased attention to auditory feedback improves its processing and its integration in speech control, which reduces stuttering. Developmental stuttering is a speech disorder that affects about one percent of the adult population worldwide. The cause is still unknown, but not a few researchers have suspected poor auditory-motor integration to be a causal factor. Almost all people who stutter are more or totally fluent in certain conditions, for instance, when speaking in chorus or alone to themselves, in time with the clicking of a metronome, or with delayed or frequency-altered auditory feedback. Understanding why stuttering disappears in these specific conditions could help understand why it occurs in normal speaking conditions. Here, the first unified account for the effect is proposed, based on evidence that receptive speech processing depends on attention to the speech signal. It is proposed that fluency-enhancing conditions re-allocate the speaker's attention, that is, the perceptual and processing capacities, by drawing attention to the auditory feedback of speech. This improves auditory-motor integration, which reduces stuttering. The hypothesis is discussed for all well-known fluency-enhancing conditions, and consequences for therapy are outlined. [ABSTRACT FROM AUTHOR]

The present investigation examined how factors such as cleft type, age of primary palatal surgery, diagnosed syndromes, hearing problems, and malocclusions could predict persistent speech difficulties and the need for speech services in school-aged children with cleft palate. Participants included 100 school-aged children with cleft palate. Americleft speech protocol was used to assess the perceptual aspects of speech production. The logistic regression was performed to evaluate the impact of independent variables (IV) on the dependent variables (DV): intelligibility, posterior oral CSCs, audible nasal emission, hypernasality, anterior oral CSCs, and speech therapy required. Sixty-five percent of the children were enrolled in (or had received) speech therapy. The logistic regression model shows a good fit to the data for the need for speech therapy (Hosmer and Lemeshow's χ 2 (8) = 9.647 , p =. 291). No IVs were found to have a significant impact on the need for speech therapy. A diagnosed syndrome was associated with poorer intelligibility (Pulkstenis-Robinson's χ 2 (11) = 7.120 , p =. 789). Children with diagnosed syndromes have about six times the odds of a higher hypernasality rating (Odds Ratio = 5.703) than others. The cleft type was significantly associated with audible nasal emission (Fisher ' s exact p =. 006). At the same time, malocclusion had a significant association with anterior oral CSCs (Fisher ' s exact p =. 005). According to the latest data in the Cleft Registry and Audit Network Annual Report for the UK, the majority of children with cleft palate attain typical speech by age five. However, it is crucial to delve into the factors that may influence the continuation of speech disorders beyond this age. This understanding is vital for formulating intervention strategies aimed at mitigating the long-term effects of speech disorders as individuals grow older. [ABSTRACT FROM AUTHOR]

• A novel deep learning model for classification of tympanic membrane conditions. • Hypercolumn deep features extracted from keypoints in otoscopy images. • Bidirectional LSTM trained by deep features in the form of non-time related data. • Obtained average accuracy of 99.06% for 4 different classes of tympanic membrane. • Proposed model can assist the otolaryngologist to make accurate diagnosis. Objectives : Middle ear inflammatory diseases are global health problem that can have serious consequences such as hearing loss and speech disorders. The high cost of medical devices such as oto-endoscope and oto-microscope used by the specialists for the diagnosis of the disease prevents its widespread use. In addition, the decisions of otolaryngologists may differ due to the subjective visual examinations. For this reason, computer-aided middle ear disease diagnosis systems are needed to eliminate subjective diagnosis and high cost problems. To this aim, a hybrid deep learning approach was proposed for automatic recognition of different tympanic membrane conditions such as earwax plug, myringosclerosis, chronic otitis media and normal from the otoscopy images. Materials and methods : In this study we used public Ear Imagery dataset containing 880 otoscopy images. The proposed approach detects keypoints from the otoscopy images and following the obtained keypoint positions, extracts hypercolumn deep features from 5 different layers of the VGG 16 model. Classification of tympanic membrane conditions were realized by feeding the deep hypercolumn features to Bi-LSTM network in the form of non-time related data. Results : The performance of the proposed model was evaluated in three different color spaces as Red-Green-Blue (RGB), Hue-Saturation-Value (HSV) and Haematoxylin-Eosin-Diaminobenzidine (HED). The proposed model achieved acceptable results in all color spaces, moreover it showed a very successful performance in classifying tympanic membrane conditions especially in RGB space. Experimental studies showed that the proposed model achieved Acc of 99.06%, Sen of 98.13% and Spe of 99.38%. Conclusion : As a result, a robust model with high sensitivity was obtained for classification of tympanic membrane conditions and it was shown that Bi-LSTM network, which is generally used with time-related data, could also be used successfully with non-time related data for diagnosis of tympanic membrane conditions. [ABSTRACT FROM AUTHOR]

• Large language models (GPT-3.5, GPT-4, and Llama3) could predict human-generated ratings of coherence, content, and tangentiality in clinical speech samples with moderate accuracy. • GPT-4 and Llama3 predictions were more accurate to the human ratings than GPT-3.5. • Using default parameters, both GPT models showed high levels of inconsistency in their predicted ratings, while Llama3 was relatively consistent. • Inconsistency in large language model predictions could be mitigated by controlling the temperature parameter. • Ensemble methods which combined multiple sets of model-predicted ratings improved predictive accuracy. Natural Language Processing (NLP) methods have shown promise for the assessment of formal thought disorder, a hallmark feature of schizophrenia in which disturbances to the structure, organization, or coherence of thought can manifest as disordered or incoherent speech. We investigated the suitability of modern Large Language Models (LLMs - e.g., GPT-3.5, GPT-4, and Llama 3) to predict expert-generated ratings for three dimensions of thought disorder (coherence, content, and tangentiality) assigned to speech samples collected from both patients with a diagnosis of schizophrenia (n = 26) and healthy control participants (n = 25). In addition to (1) evaluating the accuracy of LLM-generated ratings relative to human experts, we also (2) investigated the degree to which the LLMs produced consistent ratings across multiple trials, and we (3) sought to understand the factors that impacted the consistency of LLM-generated output. We found that machine-generated ratings of the level of thought disorder in speech matched favorably those of expert humans, and we identified a tradeoff between accuracy and consistency in LLM ratings. Unlike traditional NLP methods, LLMs were not always consistent in their predictions, but these inconsistencies could be mitigated with careful parameter selection and ensemble methods. We discuss implications for NLP-based assessment of thought disorder and provide recommendations of best practices for integrating these methods in the field of psychiatry. [ABSTRACT FROM AUTHOR]

• Investigated functional brain changes in schizophrenia-related symptoms. • Explored anxiety, depression, negative symptoms & their neural correlates. • Addressed broader clinical manifestations beyond core symptoms. • Utilized advanced ML algorithms for fMRI data analysis. • Provided insights into schizophrenia's complex cognitive profiles. Schizophrenia is a persistent neurological disorder profoundly affecting cognitive, emotional, and behavioral functions, prominently characterized by delusions, hallucinations, disordered speech, and abnormal motor activity. These symptoms often present diagnostic challenges due to their overlap with other forms of psychosis. Therefore, the implementation of automated diagnostic methodologies is imperative. This research leverages Functional Magnetic Resonance Imaging (fMRI), a neuroimaging modality capable of delineating functional activations across diverse brain regions. Furthermore, the utilization of evolving machine learning techniques for fMRI data analysis has significantly progressive. Here, our study stands as a novel attempt, focusing on the comprehensive assessment of both classical and atypical symptoms of schizophrenia. We aim to uncover associated changes in brain functional activity. Our study encompasses two distinct fMRI datasets (1.5T and 3T), each comprising 34 schizophrenia patients for the 1.5T dataset and 25 schizophrenia patients for the 3T dataset, along with an equal number of healthy controls. Machine learning algorithms are applied to assess data subsets, enabling an in-depth evaluation of the current functional condition concerning symptom impact. The identified voxels contribute to determining the brain regions most influenced by each symptom, as quantified by symptom intensity. This rigorous approach has yielded various new findings while maintaining an impressive classification accuracy rate of 97 %. By elucidating variations in activation patterns across multiple brain regions in individuals with schizophrenia, this study contributes to the understanding of functional brain changes associated with the disorder. The insights gained may inform differential clinical interventions and provide a means of assessing symptom severity accurately, offering new avenues for the management of schizophrenia. [ABSTRACT FROM AUTHOR]

• Modified a neural multi-talker TTS by adding a dysarthria severity level coefficient and a pause insertion model to synthesize dysarthric speech for varying severity levels. • Providing data augmentation for machine learning tasks such as automatic speech recognition. • Evaluating the effectiveness of the approach for dysarthria-specific speech recognition on the TORGO dataset, with results provided for two different experiments: the first includes augmented speech across 3 severities with pause insertion, and the second includes augmented speech with across a larger number of variables that include severity, pause, pitch, energy, and duration. • A relative improvement of 12.2 % on word error rate (WER), demonstrating that using dysarthric synthetic speech to increase the amount of dysarthric-patterned speech for training has the potential for significant impact on dysarthric ASR systems. • Two subjective evaluations of the synthesized dysarthric speech are provided. This includes an evaluation of Dysarthric-ness that shows that the perceived level of the dysarthria tracks with the target synthesized dysarthric severity, as well as an evaluation of speaker similarity that shows a higher rating of similarity between synthesis target speaker and actual speaker when these are the same individual. • A demonstration web page with audio results of the synthesis is available at https://mohammadelc.github.io/SpeechGroupUKY/. Dysarthria is a motor speech disorder often characterized by reduced speech intelligibility through slow, uncoordinated control of speech production muscles. Automatic Speech Recognition (ASR) systems can help dysarthric talkers communicate more effectively. However, robust dysarthria-specific ASR requires a significant amount of training speech, which is not readily available for dysarthric talkers. This paper presents a new dysarthric speech synthesis method for the purpose of ASR training data augmentation. Differences in prosodic and acoustic characteristics of dysarthric spontaneous speech at varying severity levels are important components for dysarthric speech modeling, synthesis, and augmentation. For dysarthric speech synthesis, a modified neural multi-talker TTS is implemented by adding a dysarthria severity level coefficient and a pause insertion model to synthesize dysarthric speech for varying severity levels. To evaluate the effectiveness for synthesis of training data for ASR, dysarthria-specific speech recognition was used. Results show that a DNN HMM model trained on additional synthetic dysarthric speech achieves relative Word Error Rate (WER) improvement of 12.2 % compared to the baseline, and that the addition of the severity level and pause insertion controls decrease WER by 6.5 %, showing the effectiveness of adding these parameters. Overall results on the TORGO database demonstrate that using dysarthric synthetic speech to increase the amount of dysarthric-patterned speech for training has significant impact on the dysarthric ASR systems. In addition, we have conducted a subjective evaluation to evaluate the dysarthricness and similarity of synthesized speech. Our subjective evaluation shows that the perceived dysarthricness of synthesized speech is similar to that of true dysarthric speech, especially for higher levels of dysarthria. Audio samples are available at https://mohammadelc.github.io/SpeechGroupUKY/ [ABSTRACT FROM AUTHOR]

Stuttering is a prevalent speech disorder affecting millions worldwide. To provide an automatic and objective stuttering assessment tool, Stuttering Event Detection (SED) is under extensive investigation for advanced speech research and applications. Despite significant progress achieved by various machine learning and deep learning models, SED directly from speech signal is still challenging due to stuttering speech's heterogeneous and overlapped nature. This paper presents a novel SED approach using multi-feature fusion and attention mechanisms. The model utilises multiple acoustic features extracted based on different pitch, time-domain, frequency domain, and automatic speech recognition feature to detect stuttering core behaviours more accurately and reliably. In addition, we exploit both spatial and temporal attention mechanisms as well as Bidirectional Long Short-Term Memory (BI-LSTM) modules to learn better representations to improve the SED performance. The experimental evaluation and analysis convincingly demonstrate that our proposed model surpasses the state-of-the-art models on two popular stuttering datasets, with 4% and 3% overall F1 scores, respectively. The superior results indicate the consistency of our proposed method, supported by both multi-feature and attention mechanisms in different stuttering events datasets. • We propose A novel attention-based model to learn stuttering events detection (SED). • Our multi-feature fusion approach can improve SED performance. • Our multi-feature fusion model outperforms state-of-the-art methods. • We introduced an attention map along two separate dimensions for SED. • We demonstrate the effectiveness of CBAM in focusing on the salient features. [ABSTRACT FROM AUTHOR]

We retrospectively analyzed data from 15 patients, with a normal pre-operative cognitive performance, undergoing awake surgery for left fronto-temporal low-grade glioma. We combined a pre-surgical measure (fMRI maps of motor- and language-related centers) with intra-surgical measures (MNI-registered cortical sites data obtained during intra-operative direct electrical stimulation, DES, while they performed the two most common language tasks: number counting and picture naming). Selective DES effects along the precentral gyrus/inferior frontal gyrus (and/or the connected speech articulation network) were obtained. DES of the precentral gyrus evoked the motor speech arrest, i.e., anarthria (with apparent mentalis muscle movements). We calculated the number of shared voxels between the lip-tongue and overt counting related- and silent naming-related fMRI maps and the Volumes of Interest (VOIs) obtained by merging together the MNI sites at which a given speech disturbance was observed, normalized on their mean the values (i.e., Z score). Both tongue- and lips-related movements fMRI maps maximally overlapped (Z = 1.05 and Z = 0.94 for lips and tongue vs. 0.16 and −1.003 for counting and naming) with the motor speech arrest seed. DES of the inferior frontal gyrus, pars opercularis and the rolandic operculum induced speech arrest proper (without apparent mentalis muscle movements). This area maximally overlapped with overt counting-related fMRI map (Z = −0.11 and Z = 0.09 for lips and tongue vs. 0.9 and 0.0006 for counting and naming). Interestingly, our fMRI maps indicated reduced Broca's area activity during silent speech compared to overt speech. Lastly, DES of the inferior frontal gyrus, pars opercularis and triangularis evoked variations of the output, i.e., dysarthria, a motor speech disorder occurring when patients cannot control the muscles used to produce articulated sounds (phonemes). Silent object naming-related fMRI map maximally overlapped (Z = −0.93 and Z = −1.04 for lips and tongue vs. −1.07 and 0.99 for counting and naming) with this seed. Speech disturbances evoked by DES may be thought of as selective interferences with specific recruitment of left inferior frontal gyrus and precentral cortex which are differentiable in terms of the specific interference induced. [ABSTRACT FROM AUTHOR]

• The autistic group had higher scores in some domains of social anxiety. • Higher social anxiety was not associated with increased speech disfluency. • Both groups had higher physiological arousal in narrating condition compared to the video viewing. • Physiological reactivity was not related to disfluencies in autistic persons. • Physiological reactivity was negatively associated with disfluencies in controls. The aim of this study was to examine possible associations of social anxiety (SA) and speaking-related physiological reactivity with the frequencies of a) total disfluencies, b) typical disfluencies, and c) stuttering-like disfluencies, as well as d) stuttering-severity in autistic young adults and controls. Thirty-two autistic young adults and 35 controls participated in this study. Participants were presented with video clips (viewing condition) and were then asked to talk about the videos (narrating condition). SA was measured by the self-report Social Phobia and Anxiety Inventory (SPAI). Speaking-related physiological reactivity was measured by the electrodermal activity (EDA), an index of emotional arousal. The speech samples from the narrating condition were analyzed for type and frequency of speech disfluencies and used for determining the stuttering severity. SA and speaking-related physiological reactivity were compared between the groups. Correlation between SA, physiological reactivity, disfluency frequencies, and stuttering severity were tested separately for both groups. No between-group differences were found in the overall SA, yet differences were found in SPAI subscales of social interaction, group interaction, and avoidance, as well as in agoraphobia. Both groups had higher physiological arousal in narrating condition in comparison to the video viewing condition, yet there was no between-group difference in the reactivity. No associations were found between SPAI measures, physiological reactivity, disfluency frequencies, and stuttering severity in the autistic group. In the control group, a negative association was found between physiological reactivity and total and typical disfluency frequencies. SA or speaking-related physiological reactivity were not associated with disfluency frequencies or stuttering severity in autistic persons. Negative association between physiological reactivity and disfluency frequencies found in the control group may indicate that the physiological arousal may impact the speech production process by reducing the overt disfluencies. [ABSTRACT FROM AUTHOR]

In this study, the challenges of adapting automatic speaker verification (ASV) systems to accommodate individuals with dysarthria, a speech disorder affecting intelligibility and articulation, are addressed. The scarcity of dysarthric speech data presents a significant obstacle in the development of an effective ASV system. To mitigate the detrimental effects of data paucity, an out-of-domain data augmentation approach was employed based on the observation that dysarthric speech often exhibits longer phoneme duration. Motivated by this observation, the duration of healthy speech data was modified with various stretching factors and then pooled into training, resulting in a significant reduction in the error rate. In addition to analyzing average phoneme duration, another analysis revealed that dysarthric speech contains crucial high-frequency spectral information. However, Mel-frequency cepstral coefficients (MFCC) are inherently designed to down-sample spectral information in the higher-frequency regions, and the same is true for Mel-filterbank features. To address this shortcoming, Linear-filterbank cepstral coefficients (LFCC) were used in combination with MFCC features. While MFCC effectively captures certain aspects of dysarthric speech, LFCC complements this by capturing high-frequency details essential for accurate dysarthric speaker verification. This proposed feature fusion effectively minimizes spectral information loss, further reducing error rates. To support the significance of combination of MFCC and LFCC features in an automatic speaker verification system for speakers with dysarthria, comprehensive experimentation was conducted. The fusion of MFCC and LFCC features was compared with several other front-end acoustic features, such as Mel-filterbank features, linear filterbank features, wavelet filterbank features, linear prediction cepstral coefficients (LPCC), frequency domain LPCC, and constant Q cepstral coefficients (CQCC). The approaches were evaluated using both i -vector and x -vector-based representation, comparing systems developed using MFCC and LFCC features individually and in combination. The experimental results presented in this paper demonstrate substantial improvements, with a 25.78% reduction in equal error rate (EER) for i -vector models and a 23.66% reduction in EER for x -vector models when compared to the baseline ASV system. Additionally, the effect of feature concatenation with variation in dysarthria severity levels (low, medium, and high) was studied, and the proposed approach was found to be highly effective in those cases as well. • The paper addresses dysarthria ASV challenges using MFCC and LFCC feature fusion, with an out-of-domain data augmentation technique. • A detailed analysis of dysarthric speech is also presented in this paper to demonstrate the presence of a significant amount of relevant spectral information in the high-frequency components of dysarthric speech. • A comparison of i -vectors and x -vectors-based speaker modeling is also conducted experimentally. • Comprehensive exploration for optimal feature set in dysarthria speaker verification was conducted. • A severity-wise analysis of performance of the developed ASV system is also done. [ABSTRACT FROM AUTHOR]

As the leading killer of life and health, stroke leads to limb paralysis, speech disorder, dysphagia, cognitive impairment, mental depression and other symptoms, which entail a significant financial burden to society and families. At present, physiology, clinical medicine, engineering, and materials science, advanced biomaterials standing on the foothold of these interdisciplinary disciplines provide new opportunities and possibilities for the cure of stroke. Among them, hydrogels have been endowed with more possibilities. It is well-known that hydrogels can be employed as potential biosensors, medication delivery vectors, and cell transporters or matrices in tissue engineering in tissue engineering, and outperform many traditional therapeutic drugs, surgery, and materials. Therefore, hydrogels become a popular scaffolding treatment option for stroke. Diverse synthetic hydrogels were designed according to different pathophysiological mechanisms from the recently reported literature will be thoroughly explored. The biological uses of several types of hydrogels will be highlighted, including pro-angiogenesis, pro-neurogenesis, anti-oxidation, anti-inflammation and anti-apoptosis. Finally, considerations and challenges of using hydrogels in the treatment of stroke are summarized. [Display omitted] • Hydrogels can play a role in stroke treatment by pro-angiogenesis, pro-neurogenesis, anti-oxidation, anti-inflammation and anti-apoptosis. • Hydrogel treatments have significant potential in treating ferroptosis, blood toxicity, reactive gliosis, and mitochondrial dysfunction following a stroke. • Hydrogels can provide significant benefits in individual and clinical stroke treatment. [ABSTRACT FROM AUTHOR]

Velocardiofacial syndrome, a prevalent microdeletion syndrome occurring in 1 in 2000–4000 live births, is marked by speech and language disorders, notably velopharyngeal dysfunction. This study investigates speech outcomes, nasometric and videofluoroscopic results before and after primary repair of cleft palate using the Sommerlad intravelar veloplasty (SIVV) technique within the Isfahan cleft care team for patients with velocardiofacial syndrome. Employing a quasi-experimental design, 19 participants with velocardiofacial syndrome, who underwent primary cleft palate repair by the Isfahan cleft care team, were included through convenience sampling. Perceptual and instrumental outcomes were assessed pre-and post-operatively. Statistical analysis encompassed paired t-tests and the non-parametric Wilcoxon signed-rank test (p < 0.05). The study identified no statistically significant differences between pre-and post-surgical speech outcome parameters and nasalance scores. Nonetheless, a significant distinction emerged in the velopharyngeal closure ratio based on fluoroscopic evaluation (p = 0.038). The efficacy of the SIVV technique in treating velopharyngeal dysfunction in velocardiofacial syndrome patients is inconclusive, demanding further research. Post-surgical speech outcomes are influenced by surgical technique, hypotonia, apraxia of speech, and surgery timing. Notably, an elevated velopharyngeal valve closure ratio, though anatomically indicative, does not exclusively predict surgical success. • Sommerlad intravelar veloplasty (SIVV) technique did not affect speech outcomes after Primary Repair of Cleft Palate in Velocardiofacial Syndrome. • Significant improvement in velopharyngeal closure ratio post-SIVV, doesn't exclusively predict the decreased overall size of the VP gap. • Surgical success is influenced by technique, hypotonia, apraxia of speech, and surgery timing. [ABSTRACT FROM AUTHOR]

Pathogenic variants in BRPF1 cause intellectual disability, ptosis and facial dysmorphism. Speech and language deficits have been identified as a manifestation of BRPF1- related disorder but have not been systematically characterized. We provide a comprehensive delineation of speech and language abilities in BRPF1 -related disorder and expand the phenotype. Speech and language, and health and medical history were assessed in 15 participants (male = 10, median age = 7 years 4 months) with 14 BRPF1 variants. Language disorders were common (11/12), and most had mild to moderate deficits across receptive, expressive, written, and social-pragmatic domains. Speech disorders were frequent (7/9), including phonological delay (6/9) and disorder (3/9), and childhood apraxia of speech (3/9). All those tested for cognitive abilities had a FSIQ ≥70 (4/4). Participants had vision impairment (13/15), fine (8/15) and gross motor delay (10/15) which often resolved in later childhood, infant feeding impairment (8/15), and infant hypotonia (9/15). We have implicated BRPF1- related disorder as causative for speech and language disorder, including childhood apraxia of speech. Adaptive behavior and cognition were strengths when compared to other monogenic neurodevelopmental chromatin-related disorders. The universal involvement of speech and language impairment is noteable, relative to the high degree of phenotypic variability in BRPF1- related disorder. [ABSTRACT FROM AUTHOR]

Background: Speech disorders are amongst the first symptoms to appear in Parkinson's disease (PD).Objectives: We aimed to characterize PD voice signature from the prodromal stage (isolated rapid eye movement sleep behavior disorder, iRBD) to early PD using an automated acoustic analysis and compare male and female patients. We carried out supervised learning classifications to automatically detect patients using voice only.Methods: Speech samples were acquired in 256 French speakers (117 participants with early PD, 41 with iRBD, and 98 healthy controls), with a professional quality microphone, a computer microphone and their own telephone. High-level features related to prosody, phonation, speech fluency and rhythm abilities were extracted. Group analyses were performed to determine the most discriminant features, as well as the impact of sex, vocal tasks, and microphone type. These speech features were used as inputs of a support vector machine and were combined with classifiers using low-level features.Results: PD related impairments were found in prosody, pause durations and rhythmic abilities, from the prodromal stage. These alterations were more pronounced in men than in women. Early PD detection was achieved with a balanced accuracy of 89% in males and 70% in females. Participants with iRBD were detected with a balanced accuracy of 63% (reaching 70% in the subgroup with mild motor symptoms).Conclusion: This study provides new insight in the characterization of sex-dependent early PD speech impairments, and demonstrates the valuable benefit of including automated voice analysis in future diagnostic procedures of prodromal PD. [ABSTRACT FROM AUTHOR]

Background: Speech disturbances are a recognised aspect of schizophrenia that may have potential utility as a diagnostic indicator. Recent advances in quantitative speech assessment methods have led to more reproducible and precise metrics making this possible. The current study sought firstly to characterise the speech profile of schizophrenia patients using quantitative speech measures, then examine the diagnostic utility of these measures and explore their relationship to symptoms.Methods: Speech recordings from 43 schizophrenia/schizoaffective disorder (SZ) patients and 46 healthy controls (HC) were obtained and transcribed. Cognitive and symptom measures were also administered.Results: Compared to HCs, SZ patients had higher incidences of aberrance across five types of quantitative speech variables: utterances, single words, time/speaking rate, turns and formulation errors, but not pauses. Based on two machine learning algorithms, 21 speech variables across the same five speech variable types (again not including pauses) were identified as significant classifiers for a schizophrenia diagnosis with 90-100% specificity and 80-90% sensitivity for both models. Selective relationships were also observed between these speech variables and only positive, disorganisation, excitement and formal thought disorder symptoms.Conclusions: The findings support pervasive speech impairments in schizophrenia patients relative to HCs, and the potential diagnostic utility of these speech disturbances. Continued work is needed to build the evidence base for quantitative speech assessment as a future objective diagnostic tool for schizophrenia. It holds the promise of improved diagnostic accuracy leading to increased treatment efficacy and better patient outcomes. [ABSTRACT FROM AUTHOR]

The evaluation of speech outcomes after resection and reconstruction of the oral tongue remains largely unsystematic. A cross-sectional study was performed to analyse the speech outcomes of patients who underwent curative treatment with appropriate reconstruction. Sixty-nine patients were assessed for speech intelligibility and phonetics using a validated speech intelligibility assessment tool in the local language. Volume defects were classified as class I (less than one third), II (one third to half), III (half to two-thirds), or IV (two-thirds to total glossectomy). Defect location was defined as lateral, tip, or sulcus. The χ2 test and Kruskal–Wallis test were used to test volume and location as predictors. Twenty-six patients had class I defects, 29 had class II defects, seven had class III defects, and seven had class IV defects. Twenty-two patients (31.9%) received adjuvant radiotherapy. Mean vowel, consonant, word, and paragraph intelligibility were 99.27%, 86.86%, 85.52%, and 88.72%, respectively. The incremental volume of the glossectomy defect was significantly correlated with speech intelligibility scores and phonatory alterations. In classes II and III, tip resection significantly affected interdental sounds. All patients in class III had affected alveolar and alveo-palatal sounds. The results positively corroborated the volume and location of the glossectomy defect to a classification system. [ABSTRACT FROM AUTHOR]

Introduction: Verbal fluency (VF) decline is a well-documented cognitive effect of Deep Brain Stimulation of the subthalamic nucleus (STN-DBS) in patients with Parkinson's disease (PD). This decline may be associated with disruption to left-sided frontostriatal circuitry involving the anteroventral non-motor area of the STN. While recent studies have examined the impact of lead location in relation to functional STN subdivisions on VF outcomes, results have been mixed and methods have been limited by atlas-based location mapping.Methods: Participants included 59 individuals with PD who underwent bilateral STN-DBS. Each participant's active contact location was determined in an atlas-independent fashion, relative to their individual MR-visualized STN midpoint. Multiple linear regression was used to examine lead location in each direction as a predictor of phonemic and semantic VF decline, controlling for demographic and disease variables.Results: More anterior lead locations relative to the STN midpoint in the left hemisphere predicted greater phonemic VF decline (B = -2.34, B SE = 1.08, β = -0.29, sr2 = 0.08). Lead location was not a significant predictor of semantic VF decline.Conclusion: Using an individualized atlas-independent approach, present findings suggest that more anterior stimulation of the left STN may uniquely contribute to post-DBS VF decline. This is consistent with models in which the anterior STN represents a "non-motor" functional subdivision with connections to frontal regions, e.g., the left dorsal prefrontal cortex. Future studies should investigate the effect of DBS lead trajectory on VF outcomes. [ABSTRACT FROM AUTHOR]

To analyze objective and subjective progression of speech intelligibility in oral cancer patients undergoing high-frequency speech therapy during early rehabilitation. Oral cancer patients in the Department of Maxillofacial Surgery, University Hospital of Schleswig-Holstein, Kiel, Germany, participated in the study from March 2016 to November 2017. Speech intelligibility was analyzed preoperatively (t1), post radiation (t2), and post speech therapy (t3). Objective measures were the Munich Intelligibility Profile (Online) and the Frenchay Dysarthria Assessment-2 (FDA-2). Subjective measures were the Speech Handicap Index (SHI), the speech subscale of the EORTC QLQ-C30&HN35, and the WHO-5 Index II. For nine patients with complete data, progression analyses showed a non-existent-to-low intelligibility impairment at t1 (means/SDs: e.g. FDA-2: 8.96/0.11, SHI: 17.5/15.15), increasing towards t2 (means/SDs/ p -values for difference from t1: e.g. FDA-2: 7.40/0.80/0.000, SHI: 21.7/14.24/0.213), and then decreasing towards t3, without ever reaching the initial level (means/SDs/ p -values for difference from t1: e.g. FDA-2: 8.22/0.60/0.005, SHI: 23.5/15.85/0.481). The objective changes in intelligibility were significant; the subjective changes were not. Overall, the ability to speak intelligibly after oral cancer treatment follows a typical pattern. Therefore, high-frequency speech therapy in the early rehabilitation phase might be recommendable. It might help patients to adapt to their situation after surgery, and facilitates compensating for possible functional deficits. [ABSTRACT FROM AUTHOR]

• We use transfer learning to classify speech from patients with Parkinson's and Huntington's disease, in different languages. • Experiments were conducted to evaluate transfer learning among speech from different diseases and languages. • Transfer learning improved accuracy to classify German and Czech Parkinson's patients using a base model from Spanish data. • Transfer learning schemes were accurate to improve the accuracy in a target corpus only when base models were robust enough. Patients suffering from neurodegenerative disorders such as Parkinson's or Huntington's disease exhibit speech impairments that affect their communication capabilities. The automatic assessment of the speech of the patients allows to develop computer aided tools to support the diagnosis and to evaluate the disease severity, which helps clinicians to make timely decisions about the treatment of the patients. This paper extends our previous studies about methods to classify patients with neurodegenerative diseases from speech. The proposed approach considers convolutional neural networks trained with time frequency representations and a transfer learning strategy to classify different speech impairments in patients that are native of different languages. The transfer learning schemes aim to improve the accuracy of the models when the weights of a neural network are initialized with utterances from a different corpus than the one used for the test set. The proposed methodology is evaluated with speech data from Parkinson's disease patients, who are Spanish, German, and Czech native speakers, Huntington's disease patients, who are Czech native speakers, and English native speakers affected by laryngeal impairments. We performed experiments in two scenarios: (1) transfer learning among languages, where a base model is transferred to classify patients with the same disease, but who speak a different language, and (2) transfer learning among diseases, where the base model is transferred to a corpus from patients with a different disease. The results suggest that the transfer learning schemes improve the accuracy in the target corpus only when the base model is accurate enough to transfer the knowledge to the target corpus. This behavior is observed in different scenarios of both transfer learning among languages and diseases. [ABSTRACT FROM AUTHOR]

To develop a real-time dynamic vocal tract imaging method using an accelerated spiral GRE sequence and low rank plus sparse reconstruction. Spiral k-space sampling has high data acquisition efficiency and thus is suited for real-time dynamic imaging; further acceleration can be achieved by undersampling k-space and using a model-based reconstruction. Low rank plus sparse reconstruction is a promising method with fast computation and increased robustness to global signal changes and bulk motion, as the images are decomposed into low rank and sparse terms representing different dynamic components. However, the combination with spiral scanning has not been well studied. In this study an accelerated spiral GRE sequence was developed with an optimized low rank plus sparse reconstruction and compared with L1-SPIRiT and XD-GRASP methods. The off-resonance was also corrected using a Chebyshev approximation method to reduce blurring on a frame-by-frame basis. The low rank plus sparse reconstruction method is sensitive to the weights of the low rank and sparse terms. The optimized reconstruction showed advantages over other methods with reduced aliasing and improved SNR. With the proposed method, spatial resolution of 1.3*1.3 mm2 with 150 mm field-of-view (FOV) and temporal resolution of 30 frames-per-second (fps) was achieved with good image quality. Blurring was reduced using the Chebyshev approximation method. This work studies low rank plus sparse reconstruction using the spiral trajectory and demonstrates a new method for dynamic vocal tract imaging which can benefit studies of speech disorders. [ABSTRACT FROM AUTHOR]

• Audience Response System (ARS)-based scores of acceptability are valid. • Listeners react more often to disruption of acceptability than to unintelligibility. • Panels of 6-10-year-old children provide valid and reliable scores of acceptability. • SLPs and other adults provide similar scores; children react less frequently. • ARS-based evaluations may be useful also in other areas of speech research. We explore the validity and reliability of an Audience Response Systems (ARS)-based measure of acceptability, applied to speech produced by children with speech sound disorder (SSD). We further explore how the suggested measure relates to an ARS-based measure of intelligibility. Finally, we explore potential differences between speech-language pathologists (SLPs), untrained adults, and children in their assessments. Fifty-three listeners participated in ARS-based assessments of acceptability and intelligibility: 19 SLPs, 18 untrained adults, and 16 children (aged 6-10 years). The listeners assessed speech samples collected from 14 children with SSD and 2 children with typical speech. The validity of the ARS-based acceptability measure was investigated through correlation analyses with reference to rating-based acceptability, as well as to measures of intelligibility and speech proficiency. The ARS-based acceptability measure correlated strongly with all assumed related measures. Listeners reacted more often to affected acceptability than to unintelligibility. Child listeners reacted less frequently to acceptability disruption than the SLPs and untrained adults. ARS-based assessment of acceptability provides a valid and reliable measure of acceptability. The ARS-based methodology captures an anticipated pattern where listeners react more often to disruptions of acceptability than to unintelligibility. Children appear less sensitive to traits signaling SSD than SLPs and other adults are; however, replication is required to establish this for sure. [ABSTRACT FROM AUTHOR]

Objectives: Although literature exists on using qualitative methods to generate potential attributes for a discrete choice experiment (DCE), there is little on selecting which attributes to include. We present a case study in which a best-worst scaling case 1 (BWS-1) survey was used to guide attribute selection for a DCE. The case study's context was the decision making of professionals around the choice of augmentative and alternative communication (AAC) systems for children with limited natural speech.Methods: BWS-1 survey attributes were generated from literature reviews and focus groups. DCE attributes were selected from BWS-1 attributes. The selection criteria were: include mostly important attributes; create coherent descriptions of children and AAC systems; address the project's research aims; have an appropriate respondent burden. Attributes' importance was judged using BWS-1 relative importance scores.Results: The BWS-1 survey included 19 child and 18 AAC device/system attributes and was administered to N = 93 AAC professionals. Four child and five device/system attributes were selected for the DCE, administered to N = 155 AAC professionals.Conclusions: In this case study BWS-1 results were useful in DCE attribute selection. Four recommendations are made for future studies: define selection criteria for DCE attributes a priori; consider the impact participant's perspective will have on BWS-1 and DCE results; clearly define key terminology at the start of the study and refine it as the study progresses to reflect interim findings; BWS will be useful when there is little existing stated preference work on a topic and/or qualitative work is difficult. [ABSTRACT FROM AUTHOR]

Speech sound disorders are a common communication impairment in childhood. Because speech disorders can negatively affect the lives and the development of children, clinical intervention is often recommended. To help with diagnosis and treatment, clinicians use instrumented methods such as spectrograms or ultrasound tongue imaging to analyse speech articulations. Analysis with these methods can be laborious for clinicians, therefore there is growing interest in its automation. In this paper, we investigate the contribution of ultrasound tongue imaging for the automatic detection of speech articulation errors. Our systems are trained on typically developing child speech and augmented with a database of adult speech using audio and ultrasound. Evaluation on typically developing speech indicates that pre-training on adult speech and jointly using ultrasound and audio gives the best results with an accuracy of 86.9%. To evaluate on disordered speech, we collect pronunciation scores from experienced speech and language therapists, focusing on cases of velar fronting and gliding of /r/. The scores show good inter-annotator agreement for velar fronting, but not for gliding errors. For automatic velar fronting error detection, the best results are obtained when jointly using ultrasound and audio. The best system correctly detects 86.6% of the errors identified by experienced clinicians. Out of all the segments identified as errors by the best system, 73.2% match errors identified by clinicians. Results on automatic gliding detection are harder to interpret due to poor inter-annotator agreement, but appear promising. Overall findings suggest that automatic detection of speech articulation errors has potential to be integrated into ultrasound intervention software for automatically quantifying progress during speech therapy. • Automatic error detection for speech therapy using ultrasound visual biofeedback. • Ultrasound tongue imaging is used to detect velar fronting and gliding of /r/. • Experienced therapists have high agreement when scoring velars, but not rhotics. • Models using ultrasound and audio outperform those using only ultrasound or audio. • Adult ultrasound data improves results for typically developing speech. [ABSTRACT FROM AUTHOR]

Parkinson's disease (PD) is a neuro-degenerative disease due to loss of brain cells, which produces dopamine. It is most common after Alzheimer's disease specially seen in old age people. In the earlier stage of disease, it has been noticed that most of the people suffering from speech disorder. From last two decades many studies have been conducted for the analysis of vocal tremors in PD. This study explores the combined approach of Variational Mode Decomposition (VMD) and Hilbert spectrum analysis (HSA) to investigate the voice tremor of patients with PD. A new set of features Hilbert cepstral coefficients (HCCs) are proposed in this study. Proposed features are assessed using vowels and words of PC-GITA database. The effectiveness of HCC features is utilized to perform classification, and regression analysis for PD detection. The highest average classification accuracy up to 91% and 96% is obtained with vowel /a/ and word /apto/ respectively. Further the classification accuracy up to 82% is obtained with independent dataset, when tested with the optimized model developed using PC-GITA database. In dysarthria level prediction highest correlation up to 0.82 is obtained using vowel /a/ and 0.8 with word /petaka/. The outcomes of this study indicate that the proposed articulatory features are suitable and accurate for PD assessment. [ABSTRACT FROM AUTHOR]

The specific contribution of focal damage of the prefrontal cortex and the cognitive mechanisms accounting for communicative-pragmatic disorders remains unclear. The objective of the current study was to investigate the impact of focal prefrontal cortex damage on the ability to understand indirect speech or hints and to identify the prefrontal neural mechanisms involved. We also examined the underlying cognitive mechanisms of disorders of indirect speech understanding particularly theory of mind and executive functions. Thirty patients with focal prefrontal cortex damage and 30 control subjects were compared on their performances on the Hinting task assessing pragmatic language skills, the "Faux-Pas" task and the Reading the Mind in the Eyes task assessing Theory of Mind and a battery of executive tasks. Patients were significantly impaired compared with control subjects on all these abilities. Both deficits of executive functions and theory of mind were able to predict impaired ability of patients in understanding non-literal meanings on the Hinting task. Finally, using voxel-based lesion analysis we identified a partially shared neural prefrontal network involved in all these abilities centered on the dorsomedial and ventral regions of the prefrontal cortex. [ABSTRACT FROM AUTHOR]

Hypokinetic dysarthria, a motor speech disorder characterized by reduced movement and control in the speech-related muscles, is mostly associated with Parkinson's disease. Acoustic speech features thus offer the potential for early digital biomarkers to diagnose and monitor the progression of this disease. However, the influence of language on the successful classification of healthy and dysarthric speech remains crucial. This paper explores the analysis of acoustic speech features, both established and newly proposed, in a multilingual context to support the diagnosis of PD. The study aims to identify language-independent and highly discriminative digital speech biomarkers using statistical analysis and machine learning techniques. The study analyzes thirty-three acoustic features extracted from Czech, American, Israeli, Columbian, and Italian PD patients, as well as healthy controls. The analysis employs correlation and statistical tests, descriptive statistics, and the XGBoost classifier. Feature importances and Shapley values are used to provide explanations for the classification results. The study reveals that the most discriminative features, with reduced language dependence, are those measuring the prominence of the second formant, monopitch, and the frequency of pauses during text reading. Classification accuracies range from 67% to 85%, depending on the language. This paper introduces the concept of language robustness as a desirable quality in digital speech biomarkers, ensuring consistent behaviour across languages. By leveraging this concept and employing additional metrics, the study proposes several language-independent digital speech biomarkers with high discrimination power for diagnosing PD. • Digital speech biomarkers enable objective assessment of hypokinetic dysarthria (HD). • Language affects speaker classification accuracy. • Introducing concept of features' language robustness. • Promising biomarkes: lower second formant prominence, monopitch, fewer pauses. • Text reading essential in HD assessment. [ABSTRACT FROM AUTHOR]

In the clinical practice of traditional Chinese medicine, HuangQi-DanShen (HD) is an important drug pair for the treatment of cerebral ischemia (CI). Elucidate the mechanism of HD against CI based on symptom-oriented network pharmacology coupled with comprehensive chemical characterization. UHPLC-Q-Exactive Orbitrap-MS technology was firstly used to obtain the chemical profile of HD constituents. A comprehensive strategy combining in-house library, diagnostic ions, Compound Discover software and network databases was then established to identify its chemical constitutes. Symptomatic treatment is a treatment aimed at relieving or eliminating symptoms which is often characterized as a stop-gap measure due to its inability to cure the disease fundamentally. Nevertheless, symptomatic treatment is an indispensable part of clinical practice and has an important place in medical therapeutics. Therefore, network pharmacology technique were used to elucidate molecular mechanisms from the symptoms of CI. Finally, some literatures were further mined to support our conclusions. A total of 190 ingredients were identified in HD. Symptom-oriented network pharmacology analysis indicated that compounds of HD relieved "blood" through the regulation of ADORA2A, ADORA1, PTPN11, MMP9 and EGFR, relieved "qi" via the regulation of ADORA2A, EGFR, MMP9 and CA2. The therapeutic effect of HD on "faint" was linked to PTPN11 and MMP9, while the regulation of "dyskinesia" was related to ADORA2A and EGFR, and ADORA1, PTPN11 and MMP9 were associated withe its effect on "speech disorder". ADORA1, ADORA2A and MMP9 were key to the HD component in treating "visual disturbance". The approach of symptom-oriented network pharmacology coupled with comprehensive chemical characterization proposed a further orientation for exploring the mechanisms of HD against CI. [Display omitted] • Symptom-oriented network pharmacology unveiled mechanisms of HD herb pair against CI. • A strategy for compound characterization of HD to aid network pharmacology analysis. • ADORA1, ADORA2A, PTPN11, CA2, EGFR and MMP9 were essential for HD treatment for CI. [ABSTRACT FROM AUTHOR]

• The shape of kinematic contours in neurotypical and impaired speech can considerably deviate from theoretically-driven models of articulatory behaviors. • Articulatory Phonology/Task Dynamics (AP/TD) has the potential to help in classifying patterns related to different motor impairments. • Introducing stochastic noise to AP/TD at different levels of the model can account for deviant motor contours in impaired speech (Proof of Concept) • Accommodating impaired articulatory behaviors within dynamical systems models can gain a more comprehensive understanding of motor speech disorders as well as the speech production mechanism in general. A common goal of kinematic studies on disordered speech is the identification of speech motor impairments that negatively impact speech function. Although it is well-known that the kinematic contours of speakers with speech disorders often deviate considerably from those of neurotypical speakers, systematic quantitative assessments of these impairment-related movement disturbances remain challenging. Kinematic measurement approaches are commonly grounded in models and theories that have emerged exclusively from observations of healthy speakers. However, often these models cannot accommodate the deviant articulatory behaviors of speakers with speech motor impairment. In the present paper, we address this problem. By considering noise as a factor in Articulatory Phonology/Task Dynamics (AP/TD), we can account for articulatory behaviors that are known to occur in healthy speakers (e.g., during slow speech) as well as in speakers with motor speech impairments. In a proof of concept, we descriptively compare modeled articulatory behaviors that include noise at various levels with empirical data. We view such an extension of the AP/TD as a first step towards a more comprehensive speech production model that can serve as a theoretical framework to study the speech production mechanism in healthy speakers and speakers with motor speech impairments. [ABSTRACT FROM AUTHOR]

• Forms of syndromic obesity are cases in which obesity occurs with additional phenotypes (e.g. intellectual disability, organ-specific developmental abnormalities). • We identified a novel mutation [ MYT1L (NM_015025.4):c.2990C>A] that causes syndromic obesity. • The phenotypic spectrum of patients with MYT1L SNVs/indels variants always includes intellectual disability/developmental delay and speech delay; most patients have polyphagia and obesity/overweight. Macrocephaly, strabismus, behavioral problems, and hands/feet malformations are common phenotypes. • The mechanism by which MYT1L haploinsufficiency results in polyphagia is possibly originated by alteration on the neuroendocrine hypothalamus development. Pathogenic variants involving the MYT1L gene lead to an autosomal dominant form of syndromic obesity, characterized by polyphagia, intellectual disability/developmental delay, and behavioral problems, and that a characteristic facial phenotype does not seem to be recognizable. Trio whole exome sequencing was performed in a 10-year-old Brazilian male presenting polyphagia, severe early-onset obesity, intellectual disability, speech delay, macrocephaly, frontal bossing, telecanthus, strabismus, and hypogenitalism. Additionally, we performed a literature review of patients carrying non-copy number MYT1L variants. A de novo genetic variant not previously reported in MYT1L (NM_015025.4:c.2990C > A) was identified in the proband and classified as pathogenic. From a literature search, 22 further patients carrying non-copy number MYT1L variants were identified, evidencing that although the associated phenotype is quite variable, intellectual disability/developmental and speech delays are always present. Further, most patients have obesity or overweight due to polyphagia. Macrocephaly, strabismus, behavioral problems, and hand/feet malformations are also recurrent features. We described the first Brazilian case of MYT1L related syndrome and highlighted clinical characteristics based on the literature. Other syndromic forms of obesity such as Prader-Willi, Bardet-Biedl, Börjeson-Forssman-Lehmann, MORM, Cohen, Alstrom, and Kleefstra type 1 syndromes should be considered in the differential diagnosis. Further, although obesity is frequent, it is not an obligatory feature of all carriers of MYT1L mutations. [ABSTRACT FROM AUTHOR]

Communicating emotion is essential in building and maintaining relationships. We communicate our emotional state not just with the words we use, but also how we say them. Changes in the rate of speech, short-term energy and intonation all help to convey emotional states like 'angry', 'sad' and 'happy'. People with dysarthria, the most common speech disorder, have reduced articulatory and phonatory control. This can affect the intelligibility of their speech, especially when communicating with unfamiliar conversation partners. However, we know little about how people with dysarthria convey their emotional state, and whether they are having to make changes to their speech to achieve this. In this study, we investigated the ability of people with dysarthria, caused by cerebral palsy and Parkinson's disease, to communicate emotions in their speech, and we compared their speech to that of speakers with typical speech. A parallel database of emotional speech was collected. One female speaker with dysarthria due to cerebral palsy, 3 speakers with dysarthria due to Parkinson's disease (2 female and 1 male), and 21 typical speakers (9 female and 12 male) produced sentences with 'angry', 'happy', 'sad', and 'neutral' emotions. A number of acoustic features were analysed using linear multi-level modeling. The results show that people with dysarthria were able to control some aspects of the suprasegmental and prosodic features when attempting to communicate emotions. For most speakers the changes they made are consistent with the changes made by speakers with typical speech. Even when the changes might be different to that of typical speakers, acoustic analysis shows these were consistent for different emotions. The analysis shows that variation in energy and jitter (local absolute) are major indicators of emotion in the study. • Acoustic analysis shows the changes made by speakers with dysarthria and typical speakers to convey their emotional states. • People with dysarthria make consistent and reliable changes to convey their emotions in a similar way to speakers with typical speech. • Variation in short term energy and jitter mark different emotions. [ABSTRACT FROM AUTHOR]

Functional movement disorders are commonly seen in neurology services and may coexist with other neurological diseases. This combination is known as "functional overlay" and an increasing interest on this topic has emerged in the past decade as the field of functional neurological disorders has moved forward. Some neurological diseases may be more prone to develop "functional overlay" than others, and within the field of movement disorders, most studies have focused on patients with Parkinson's disease. This review comprehensively summarizes the current body of knowledge on this topic and provides an expert opinion to equip clinicians with a pragmatic approach to recognize functional movement disorders in patients with Parkinson's disease, to communicate the diagnosis and to become familiar with potential therapies in this complex clinical scenario. Potential underlying mechanisms and risk factors that may play a role in increasing the vulnerability of Parkinson's disease patients to develop functional movement disorder comorbidity are also discussed within the framework of modern neurobiological theories of brain functioning. [ABSTRACT FROM AUTHOR]

This study aimed at 1) adapting the well-established Speech Handicap Index (SHI) to German, 2) testing the suitability of the instrument for assessing speech-related quality of life, 3) comparing it to the German Voice-Handicap-Index (VHI), in order to support treatment of oral cancer patients who experience posttreatment speech difficulties that affect their quality of life. Participants completed a web-based survey that employed a 2 (experienced problem: speech/articulation-related vs. voice-related) x 2 (SHI vs. VHI) between-subject experimental design, enabling it to distinguish between the experiences of voice and intelligibility impairments, and to determine the discriminatory ability of the two instruments. The German SHI reliably assessed speech intelligibility and articulation-related Quality of life. While voice impairments were equally well assessed by both, VHI: M 2.48, SD 0.65; SHI: M 2.52, SD 0.63; only the latter appropriately registered intelligibility handicap in speech impairments (VHI: M 2.05, SD 0.70; SHI: 2.68, SD 0.73). The responsivity of the SHI in capturing the experienced handicap was significantly greater in the speech/articulation-impairment condition (p =.001). The German SHI is a reliable and responsive measure for speech intelligibility and articulation-related quality of life that should be chosen in preference to the VHI. [ABSTRACT FROM AUTHOR]

• Speech auditory-motor adaptation is limited in both children who stutter (CWS) and adults who stutter (AWS). • CWS' speech adaptation is minimal but correlated with age, suggesting a protracted development of auditory-motor learning. • Reach visuomotor adaptation is subtly different in AWS but not CWS as even nonstuttering children show limited learning. Stuttering is a neurodevelopmental disorder of speech fluency. Various experimental paradigms have demonstrated that affected individuals show limitations in sensorimotor control and learning. However, controversy exists regarding two core aspects of this perspective. First, it has been claimed that sensorimotor learning limitations are detectable only in adults who stutter (after years of coping with the disorder) but not during childhood close to the onset of stuttering. Second, it remains unclear whether stuttering individuals' sensorimotor learning limitations affect only speech movements or also unrelated effector systems involved in nonspeech movements. We report data from separate experiments investigating speech auditory-motor learning (N = 60) and limb visuomotor learning (N = 84) in both children and adults who stutter versus matched nonstuttering individuals. Both children and adults who stutter showed statistically significant limitations in speech auditory-motor adaptation with formant-shifted feedback. This limitation was more profound in children than in adults and in younger children versus older children. Between-group differences in the adaptation of reach movements performed with rotated visual feedback were subtle but statistically significant for adults. In children, even the nonstuttering groups showed limited visuomotor adaptation just like their stuttering peers. We conclude that sensorimotor learning is impaired in individuals who stutter, and that the ability for speech auditory-motor learning—which was already adult-like in 3–6 year-old typically developing children—is severely compromised in young children near the onset of stuttering. Thus, motor learning limitations may play an important role in the fundamental mechanisms contributing to the onset of this speech disorder. [ABSTRACT FROM AUTHOR]

• Rare variants disrupting speech, in genes like FOXP2 , give insight into neurobiology of key human traits. • Neural functions of FOXP2 are being studied in human cell culture and animal models. • Beyond FOXP2 , next-generation sequencing implicates novel genes in speech disorders. • Some genes harboring rare variants that disrupt speech are also linked to broader syndromes. • Convergent findings implicate regulatory genes co-expressed in developing human brain. Rare genetic variants that disrupt speech development provide entry points for deciphering the neurobiological foundations of key human capacities. The value of this approach is illustrated by FOXP2 , a transcription factor gene that was implicated in speech apraxia, and subsequently investigated using human cell-based systems and animal models. Advances in next-generation sequencing, coupled to de novo paradigms, facilitated discovery of etiological variants in additional genes in speech disorder cohorts. As for other neurodevelopmental syndromes, gene-driven studies show blurring of boundaries between diagnostic categories, with some risk genes shared across speech disorders, intellectual disability and autism. Convergent evidence hints at involvement of regulatory genes co-expressed in early human brain development, suggesting that etiological pathways could be amenable for investigation in emerging neural models such as cerebral organoids. [ABSTRACT FROM AUTHOR]

Introduction: Primary progressive apraxia of speech (PPAOS) is a neurodegenerative syndrome in which patients present with an isolated motor speech disorder. Some PPAOS patients develop parkinsonism and other features of progressive supranuclear palsy (PSP) and/or corticobasal syndrome (CBS) over time. We aimed to assess the evolution of parkinsonian characteristics in PPAOS patients who had been followed yearly for at least six years.Methods: From a large cohort of 46 PPAOS patients, eight were followed yearly for > 6-years in multiple NIH-funded grants. Parkinsonian and other features, including bradykinesia, tremor, rigidity, postural instability, apraxia, ocular motor function and cognition were assessed at each visit, and research criteria applied for PSP and CBS diagnosis. Neurological, speech-language test scores, and [18F]fluorodeoxyglucose PET (FDG-PET) and MRI midbrain volumes were assessed.Results: A Parkinson's plus syndrome developed in all eight patients (100%). Bradykinesia was the earliest feature, followed by rigidity and postural instability. Tremor was not a significant feature. Parkinsonism, limb apraxia and ocular motor impairment tended to develop four-to-five years after onset with some patients having slight asymmetric parkinsonism. Six patients (75%) met research criteria for probable PSP, although only one for PSP-Richardson's syndrome; three patients met criteria for possible CBS. Slightly asymmetric, left-sided, hypometabolism was observed on FDG-PET, not matching asymmetry of Parkinsonism. Midbrain hypometabolism was absent-minimal. Three patients had progressive midbrain volumes in the PSP-Richardson's syndrome range.Conclusions: A Parkinson's plus syndrome may inevitably develop in PPAOS supporting PPAOS as an early presentation of a Parkinson's plus disorder. [ABSTRACT FROM AUTHOR]

Dysarthria is a motor speech disorder leading to imprecise articulation of speech. Acoustic analysis capable of detecting and assessing articulation errors is useful in dysarthria diagnosis and therapy. Since speakers with dysarthria experience difficulty in producing rhotics due to complex articulatory gestures of these sounds, the hypothesis of the present study is that duration of the rhotic approximant /ɹ/ distinguishes dysarthric speech of different severity levels. Duration measurements were conducted using the third formant (F3) trajectories estimated from quasi-closed-phase (QCP) spectrograms. Results indicate that the severity level of spastic dysarthria has a significant effect on duration of /ɹ/. In addition, the phonetic context has a significant effect on duration of /ɹ/, the ɪ-r-ɛ context showing the largest difference in /ɹ/ duration between dysarthric speech of the highest severity levels and healthy speech. The results of this preliminary study can be used in the future to develop signal processing and machine learning methods to automatically predict the severity level of spastic dysarthria from speech signals. • Utilization of duration of rhotic approximant /ɹ/ to assess dysarthric speech. • Utilization of QCP spectrograms in measuring the duration of rhotic approximant. • This study reveals the effect of dysarthria severity on the duration of /ɹ/. • In addition, the study shows the effect of phonetic context on /ɹ/ duration. [ABSTRACT FROM AUTHOR]

• A two-stage framework is proposed for the automatic assessment of intelligibility of dysarthric speech using glottal features. • In the first stage, two-class severity classification of dysarthria is performed using support vector machines (SVMs). • In the second stagem, intelligibility estimation of dysarthric speech is computed using a linear regression model. • Two sets of glottal parameters are explored: (1) time-domain and frequency-domain parameters and (2) parameters based on principal component analysis (PCA). • Evaluation results show improved intelligibility assessment measures (correlation and root mean square error) for the glottal parameters compared to the baseline features. Objective intelligibility assessment of dysarthric speech can assist clinicians in diagnosis of speech disorders as well as in medical treatment. This study investigates the use of glottal parameters (i.e. parameters that describe the acoustical excitation of voiced speech, the glottal flow) in the automatic intelligibility assessment of dysarthric speech. Instead of directly predicting the intelligibility of dysarthric speech using a single-stage system, the proposed method utilizes a two-stage framework. In the first stage, two-class severity classification of dysarthria is performed using support vector machines (SVMs). In the second stage, intelligibility estimation of dysarthric speech is computed using a linear regression model. Two sets of glottal parameters are explored: (1) time-domain and frequency-domain parameters and (2) parameters based on principal component analysis (PCA). Acoustic parameters proposed in a similar intelligibility prediction study by Falk et al. (2012) are used as baseline features. Evaluation results show that the two-stage framework leads to improvement in the intelligibility assessment measures (correlation and root mean square error) compared to the single-stage framework. The combination of the glottal parameters sets results in better performance in the severity classification and intelligibility estimation tasks compared to the baseline features. [ABSTRACT FROM AUTHOR]

• A representation learning scheme based on convolutional and recurrent autoencoders is proposed to model pathological speech signals with morphological and neurological origins such as cleft lip and palate and Parkinson's disease. • Two features sets are extracted from the autoencoders to model pathological speech: the hidden representation in the bottleneck space, and a proposed feature set based on the reconstruction error in different spectral components of the speech signal. • The reconstruction error from the autoencoders in different frequency regions contain information related to specific speech symptoms of both diseases. • The proposed models are accurate to classify the speech signals of both kinds of diseases, with an accuracy of up to 97% for cleft lip and palate, and up to 84% for the case of Parkinson's disease. • The results obtained in this study are comparable to those observed in the literature when the same data is considered, both for classification of cleft lip and palate, and Parkinson's disease. Speech signals may contain different paralinguistic aspects such as the presence of pathologies that affect the proper communication capabilities of a speaker. Those speech disorders have different origin depending on the type of the disease. For instance, diseases with morphological origin such as cleft lip and palate that causes hypernasality, or with neurodegenerative origin such as Parkinson's disease that generates hypokinetic dysarthria on the patients. Automatic assessment of pathological speech allows to support the diagnosis and/or the evaluation of the disease severity. Conventional methods are based on the manually applied assessment of single features such as jitter, shimmer, or formant frequencies that may not completely model all of the phenomena that appear due to the disease. This paper introduces a novel strategy based on unsupervised representation learning for automatic detection of pathological speech. The proposed approach is based on the use of recurrent and convolutional autoencoders trained to extract informative features to characterize the presence of pathologies in speech. A novel feature set based on the reconstruction error of the autoencoders is also proposed. The performance of the introduced models is evaluated classifying pathological speech signals recorded from people suffering from Parkinson's disease, and children with cleft lip and palate. All participants from this study were Spanish native speakers. The proposed models are accurate to classify the speech signals of both kinds of diseases, with an accuracy of up to 97% for cleft lip and palate, and up to 84% for the case of Parkinson's disease. We also show that the reconstruction error from the autoencoders in different frequency regions contain information related to specific speech symptoms of both diseases. [ABSTRACT FROM AUTHOR]