Speech Task Research Articles

Digital assessment of speech in Huntington disease.

Speech changes are an early symptom of Huntington disease (HD) and may occur prior to other motor and cognitive symptoms. Assessment of HD commonly uses clinician-rated outcome measures, which can be limited by observer variability and episodic administration. Speech symptoms are well suited for evaluation by digital measures which can enable sensitive, frequent, passive, and remote administration. We collected audio recordings using an external microphone of 36 (18 HD, 7 prodromal HD, and 11 control) participants completing passage reading, counting forward, and counting backwards speech tasks. Motor and cognitive assessments were also administered. Features including pausing, pitch, and accuracy were automatically extracted from recordings using the BioDigit Speech software and compared between the three groups. Speech features were also analyzed by the Unified Huntington Disease Rating Scale (UHDRS) dysarthria score. Random forest machine learning models were implemented to predict clinical status and clinical scores from speech features. Significant differences in pausing, intelligibility, and accuracy features were observed between HD, prodromal HD, and control groups for the passage reading task (e.g., p < 0.001 with Cohen'd = -2 between HD and control groups for pause ratio). A few parameters were significantly different between the HD and control groups for the counting forward and backwards speech tasks. A random forest classifier predicted clinical status from speech tasks with a balanced accuracy of 73% and an AUC of 0.92. Random forest regressors predicted clinical outcomes from speech features with mean absolute error ranging from 2.43-9.64 for UHDRS total functional capacity, motor and dysarthria scores, and explained variance ranging from 14 to 65%. Montreal Cognitive Assessment scores were predicted with mean absolute error of 2.3 and explained variance of 30%. Speech data have the potential to be a valuable digital measure of HD progression, and can also enable remote, frequent disease assessment in prodromal HD and HD. Clinical status and disease severity were predicted from extracted speech features using random forest machine learning models. Speech measurements could be leveraged as sensitive marker of clinical onset and disease progression in future clinical trials.

Isolating Neural Signatures of Conscious Speech Perception with a No-Report Sine-Wave Speech Paradigm.

Identifying neural correlates of conscious perception is a fundamental endeavor of cognitive neuroscience. Most studies so far have focused on visual awareness along with trial-by-trial reports of task-relevant stimuli, which can confound neural measures of perceptual awareness with postperceptual processing. Here, we used a three-phase sine-wave speech paradigm that dissociated between conscious speech perception and task relevance while recording EEG in humans of both sexes. Compared with tokens perceived as noise, physically identical sine-wave speech tokens that were perceived as speech elicited a left-lateralized, near-vertex negativity, which we interpret as a phonological version of a perceptual awareness negativity. This response appeared between 200 and 300 ms after token onset and was not present for frequency-flipped control tokens that were never perceived as speech. In contrast, the P3b elicited by task-irrelevant tokens did not significantly differ when the tokens were perceived as speech versus noise and was only enhanced for tokens that were both perceived as speech and relevant to the task. Our results extend the findings from previous studies on visual awareness and speech perception and suggest that correlates of conscious perception, across types of conscious content, are most likely to be found in midlatency negative-going brain responses in content-specific sensory areas.

The combination of accent method and phonemic contrast: an innovative strategy to improve speech production on post-stroke dysarthria.

Speech production includes segmental and suprasegmental features, which interact and cooperate with each other. Impaired speech production is common in individuals with post-stroke dysarthria. The commonly used phonemic contrast therapy and Accent Method in clinical practice can improve the segmental and suprasegmental aspects, respectively. This study aimed to explore the feasibility and immediate effectiveness of the combination of specific rhythm patterns of the Accent Method and phonemic contrast on speech production. Fifteen poststroke dysarthria patients (12 males and three females) first received accentuation task of three rhythm patterns (Largo-slow, Andante-medium, Allegro-fast) and later received speech task in appropriate rhythm patterns combined with phonemic contrast materials and non-phonemic contrast materials. Speech parameters were analyzed by considering speech clarity and prosody. The results showed that the number of correct target syllables, sentence clarity, and standard deviation of intensity increased significantly, and the average length of pauses and abnormal pause times decreased significantly in Andante (medium) compared to other rhythms. The number of correct target syllables, sentence clarity, and standard deviation of intensity increased significantly compared with those in the non-phonemic contrast in Andante (medium). The combination of phonemic contrast and Accent Method was verified to have an immediate effect on speech production in Mandarin speakers with post-stroke dysarthria and could be further validated in other diseases with impaired speech production in the clinic in the future.

The influence of listener experience, measurement scale and speech task on the reliability of auditory-perceptual evaluation of vocal quality.

To assess the influence of the listener experience, measurement scales and the type of speech task on the auditory-perceptual evaluation of the overall severity (OS) of voice deviation and the predominant type of voice (rough, breathy or strain). 22 listeners, divided into four groups participated in the study: speech-language pathologist specialized in voice (SLP-V), SLP non specialized in voice (SLP-NV), graduate students with auditory-perceptual analysis training (GS-T), and graduate students without auditory-perceptual analysis training (GS-U). The subjects rated the OS of voice deviation and the predominant type of voice of 44 voices by visual analog scale (VAS) and the numerical scale (score "G" from GRBAS), corresponding to six speech tasks such as sustained vowel /a/ and /ɛ/, sentences, number counting, running speech, and all five previous tasks together. Sentences obtained the best interrater reliability in each group, using both VAS and GRBAS. SLP-NV group demonstrated the best interrater reliability in OS judgment in different speech tasks using VAS or GRBAS. Sustained vowel (/a/ and /ɛ/) and running speech obtained the best interrater reliability among the groups of listeners in judging the predominant vocal quality. GS-T group got the best result of interrater reliability in judging the predominant vocal quality. The time of experience in the auditory-perceptual judgment of the voice, the type of training to which they were submitted, and the type of speech task influence the reliability of the auditory-perceptual evaluation of vocal quality.

Within-talker and within-session stability of acoustic characteristics of conversational and clear speaking stylesa).

In speech production research, talkers often perform a speech task several times per recording session with different speaking styles or in different environments. For example, Lombard speech studies typically have talkers speak in several different noise conditions. However, it is unknown to what degree simple repetition of a speech task affects speech acoustic characteristics or whether repetition effects might offset or exaggerate effects of speaking style or environment. The present study assessed speech acoustic changes over four within-session repetitions of a speech production taskset performed with two speaking styles recorded in separate sessions: conversational and clear speech. In each style, ten talkers performed a set of three speech tasks four times. Speaking rate, median fundamental frequency, fundamental frequency range, and mid-frequency spectral energy for read sentences were measured and compared across test blocks both within-session and between the two styles. Results indicate that statistically significant changes can occur from one repetition of a speech task to the next, even with a brief practice set and especially in the conversational style. While these changes were smaller than speaking style differences, these findings support using a complete speech set for training while talkers acclimate to the task and to the laboratory environment.

Detecting fatigue in multiple sclerosis through automatic speech analysis.

Multiple sclerosis (MS) is a chronic neuroinflammatory disease characterized by central nervous system demyelination and axonal degeneration. Fatigue affects a major portion of MS patients, significantly impairing their daily activities and quality of life. Despite its prevalence, the mechanisms underlying fatigue in MS are poorly understood, and measuring fatigue remains a challenging task. This study evaluates the efficacy of automated speech analysis in detecting fatigue in MS patients. MS patients underwent a detailed clinical assessment and performed a comprehensive speech protocol. Using features from three different free speech tasks and a proprietary cognition score, our support vector machine model achieved an AUC on the ROC of 0.74 in detecting fatigue. Using only free speech features evoked from a picture description task we obtained an AUC of 0.68. This indicates that specific free speech patterns can be useful in detecting fatigue. Moreover, cognitive fatigue was significantly associated with lower speech ratio in free speech (ρ = -0.283, p = 0.001), suggesting that it may represent a specific marker of fatigue in MS patients. Together, our results show that automated speech analysis, of a single narrative free speech task, offers an objective, ecologically valid and low-burden method for fatigue assessment. Speech analysis tools offer promising potential applications in clinical practice for improving disease monitoring and management.

Speech and lexico-semantic errors during direct cortical stimulation mapping of the language-dominant hemisphere: effects of object and action naming.

In this retrospective study, the authors aimed to establish the stereotactically defined probability distribution for speech (i.e., anarthria, speech arrest) and lexico-semantic errors (i.e., anomia) through direct cortical stimulation (DCS) by using two tasks: action naming and object naming. They also analyzed the patterns of interindividual variability in the localization of the language sites involved, and investigated whether any patient or lesion location factors were associated with greater variability. Eighty-one Italian-speaking patients who underwent awake surgery between 2010 and 2021 for low- and high-grade gliomas in eloquent areas of the language-dominant hemisphere were entered in the analyses. The intraoperative DCS protocol included automatic speech tasks, object naming, and action naming. The position of the tags, as depicted on the intraoperative video or photograph, was transposed into Montreal Neurological Institute space. Subsequently, a 2D scatterplot and cluster analysis were performed. Associations between various clinical and radiological characteristics and the quantity of positive stimulated sites were determined by univariate analyses using binary logistic regression. Associated variables (p < 0.2) were included in stepwise multivariate logistic regression with backward elimination (p < 0.05). A total of 1380 cortical sites were stimulated, with a positive response in 511 cases (37%). Most anarthric errors were triggered when stimulating the left precentral gyrus, and most speech arrest errors were elicited when stimulating the left posterior inferior frontal gyrus. Anomias were found in the left inferior frontal gyrus and in the posterior part of the left temporal lobe for object naming. DCS to the left dorsal premotor cortex elicited anomic errors for action naming. Anomias were also elicited during DCS to the left posterior temporal lobe, with both object and action naming. The distribution of speech and lexico-semantic errors is in line with the current literature. The action-naming results are new and mostly involve the dorsal premotor cortex. These findings stress the importance of maximizing the use of different language tasks during surgery, because even when looking for the same type of errors, different tasks may be better suited to map specific brain regions. DCS with action and object naming identifies more positive sites than object naming alone.

"A Cross-linguistic Analysis of Discourse Marker Use in Different Speech Tasks"

"A Cross-linguistic Analysis of Discourse Marker Use in Different Speech Tasks"

Research on the Innovation of Business English Teaching Mode in Colleges and Universities under the Background of Artificial Intelligence

Abstract This paper examines the application of artificial intelligence technology in business English teaching and artificial intelligence in natural language processing and recognition to receive good results. These two technologies are suitable for business English teaching innovation. At this time, the introduction of the concept of ADDIE is the process of integrating artificial intelligence technology into the business English teaching innovation model. The business English teaching model is designed to utilize the conditional random field expansion model to address the label bias limitations of the entropy model. Add the LSTM model to this basis to handle speech recognition and translation tasks to complete the construction of a business English teaching model based on LSTM-CRF. After speech recognition, detect the existence of grammatical errors in the sentence through the multi-task model, deform the LSTM model, and use the deformed Bi-LSTM to construct the multi-task learning model framework. The objective of controlled experiments and empirical analyses is to investigate the students’ willingness to learn AI Business English and their learning effects. The experiment shows that among the seven items of students’ willingness survey, only the mean value of the hedonic motivation variable is 3.93, and the mean value of the remaining six items is above 4. Students’ willingness to learn AI is high. In the comprehensive learning engagement degree, the mean value of the engagement is 1.457, and the highest value reaches 1.83278, and the students have a high degree of engagement in AI English learning.

A User-Centered Framework to Empower People with Parkinson's Disease

We present a user-centric validation of a teleneurology platform, assessing its effectiveness in conveying screening information, facilitating user queries, and offering resources to enhance user empowerment. This validation process is implemented in the setting of Parkinson's disease (PD), in collaboration with a neurology department of a major medical center in the USA. Our intention is that with this platform, anyone globally with a webcam and microphone-equipped computer can carry out a series of speech, motor, and facial mimicry tasks. Our validation method demonstrates to users a mock PD risk assessment and provides access to relevant resources, including a chatbot driven by GPT, locations of local neurologists, and actionable and scientifically-backed PD prevention and management recommendations. We share findings from 91 participants (48 with PD, 43 without) aimed at evaluating the user experience and collecting feedback. Our framework was rated positively by 80.85% (standard deviation ± 8.92%) of the participants, and it achieved an above-average 70.42 (standard deviation ± 13.85) System-Usability-Scale (SUS) score. We also conducted a thematic analysis of open-ended feedback to further inform our future work. When given the option to ask any questions to the chatbot, participants typically asked for information about neurologists, screening results, and the community support group. We also provide a roadmap of how the knowledge generated in this paper can be generalized to screening frameworks for other diseases through designing appropriate recording environments, appropriate tasks, and tailored user-interfaces.

A positive side effect of wearing face coverings for socially anxious females: Findings from a speech task

BackgroundWearing face masks has become more common due to the COVID-19 pandemic. Until now, it has not been investigated whether socially anxious individuals can reduce their acute anxiety symptoms by wearing face masks during a speech task. MethodFifty-nine socially anxious females were asked to prepare and give an oral presentation. Participants were randomly allocated either to a group that was asked to wear a face mask during the task, or to a group that was not asked to wear a face mask during the task. Dependent variables included physiological parameters (systolic/diastolic blood pressure, heart rate) and self-reports (valence and arousal at baseline, as well as directly before and after the presentation). ResultsThe analyses showed that the group without face masks had a higher systolic blood pressure than the group with face masks, directly before as well as after the presentation. The two groups did not differ in the other measures. ConclusionWearing face masks has a small stress-reducing effect on socially anxious females.

Speech and beatboxing cooperate and compromise in beatrhyming

Beatrhyming is a type of multi-vocalism in which an individual beatboxes and speaks (i.e., sings or raps) at the same time by interweaving beatboxing sounds and speech sounds within words and phrases. The measurements in this case study of a beatrhyming performance focus on one-to-one sound “replacements” in which a beatboxing sound is integrated into a word by taking the place of a speech sound. The analysis unfolds in two parts: first, a count of how many times beatboxing sounds used in place of speech sounds matched the intended speech sounds for vocal tract constrictor and constriction degree; and second, an assessment of whether the beatboxing sound patterns in beatrhyming (beatboxing with simultaneous lyrics) have the same degree of rhythmic structure as the beatboxing sound patterns in beatboxing (without lyrics). Despite having disparate aims, the separate speech and beatboxing systems work together to create a well-organized combined behavior. Speech tasks (i.e., communicating the linguistic message of the lyrics) are achieved in beatrhyming by replacing some speech sounds with beatboxing sounds that match the speech segment in vocal tract constrictor and in manner/constriction degree. Beatboxing tasks (i.e., establishing a musical rhythm) are achieved through the inviolable use of Outward K Snares {K} on the backbeat. Achieving both of these aims in the same performance requires flexibility and compromise between the speech and beatboxing systems. In addition to providing the first scientific description and analysis of beatrhyming, this article shows how beatrhyming offers new insight for phonological theories built to describe spoken language.

Ecologically valid speech collection in behavioral research: The Ghent Semi-spontaneous Speech Paradigm (GSSP)

This paper introduces the Ghent Semi-spontaneous Speech Paradigm (GSSP), a new method for collecting unscripted speech data for affective-behavioral research in both experimental and real-world settings through the description of peer-rated pictures with a consistent affective load. The GSSP was designed to meet five criteria: (1) allow flexible speech recording durations, (2) provide a straightforward and non-interfering task, (3) allow for experimental control, (4) favor spontaneous speech for its prosodic richness, and (5) require minimal human interference to enable scalability. The validity of the GSSP was evaluated through an online task, in which this paradigm was implemented alongside a fixed-text read-aloud task. The results indicate that participants were able to describe images with an adequate duration, and acoustic analysis demonstrated a trend for most features in line with the targeted speech styles (i.e., unscripted spontaneous speech versus scripted read-aloud speech). A speech style classification model using acoustic features achieved a balanced accuracy of 83% on within-dataset validation, indicating separability between the GSSP and read-aloud speech task. Furthermore, when validating this model on an external dataset that contains interview and read-aloud speech, a balanced accuracy score of 70% is obtained, indicating an acoustic correspondence between the GSSP speech and spontaneous interviewee speech. The GSSP is of special interest for behavioral and speech researchers looking to capture spontaneous speech, both in longitudinal ambulatory behavioral studies and laboratory studies. To facilitate future research on speech styles, acoustics, and affective states, the task implementation code, the collected dataset, and analysis notebooks are available.

Combining working memory and speech data in a digital app for early detection of cognitive impairment in primary care

AbstractBackgroundCognitive screening to detect mild cognitive impairment (MCI) and dementia in primary care settings has proven to be a challenging task. The ideal solution would be a brief, low‐burden assessment tool appropriate for use with individuals from diverse educational and cultural backgrounds that requires limited time and expertise from primary care providers. The purpose of this project was to develop a digital cognitive screening app derived from widely used cognitive and speech tasks using machine learning techniques.MethodParticipants were 104 old adults, 53 who were cognitively normal (CN) and 51 who were cognitively impaired (CI). All participants completed a working memory task (WM), followed by four speech‐language tasks, followed by a second administration of WM (WM2) to investigate the added utility of practice effects. Participants also completed the Quick Mild Cognitive Impairment (Qmci) screen. Bayesian adaptive regression trees were used to test 11 models. All models contained sex, age, education, and marital status. One model contained WM and the WM2‐WM difference. Four models contained 26 acoustic and 34 linguistic variables derived from each of the four speech tasks, and four more models were identical, except with the two WM variables added. For comparison, two models with Qmci were fitted: one with and one without the two WM variables.ResultThree models achieved cross‐validated classification accuracy of c≥0.90: WM by itself (estimated c = 0.91; Figure 1), WM and a personal narrative task (c = 0.94; Figure 2), and WM and a counting backwards task (c = 0.90). Models with picture description and procedural discourse tasks performed the worst. The QMCI‐only model yielded c = 0.91.ConclusionCombining WM, acoustic and linguistic variables involved in recounting personal memories, and the ability to benefit from prior exposure to a task discriminated CN from CI groups with a high degree of accuracy. This unique combination assesses key cognitive abilities known to be affected early in neurodegenerative diseases, executive functioning and memory, and includes both verbal and nonverbal tasks. Importantly, individuals with varying levels of cognitive functioning, including mild dementia, were able to use the app successfully, and the automated nature of the tool required minimal staff time or expertise.

Combining working memory and speech data in a digital app for early detection of cognitive impairment in primary care

AbstractBackgroundCognitive screening to detect mild cognitive impairment (MCI) and dementia in primary care settings has proven to be a challenging task. The ideal solution would be a brief, low‐burden assessment tool appropriate for use with individuals from diverse educational and cultural backgrounds that requires limited time and expertise from primary care providers. The purpose of this project was to develop a digital cognitive screening app derived from widely used cognitive and speech tasks using machine learning techniques.MethodParticipants were 104 old adults, 53 who were cognitively normal (CN) and 51 who were cognitively impaired (CI). All participants completed a working memory task (WM), followed by four speech‐language tasks, followed by a second administration of WM (WM2) to investigate the added utility of practice effects. Participants also completed the Quick Mild Cognitive Impairment (Qmci) screen. Bayesian adaptive regression trees were used to test 11 models. All models contained sex, age, education, and marital status. One model contained WM and the WM2‐WM difference. Four models contained 26 acoustic and 34 linguistic variables derived from each of the four speech tasks, and four more models were identical, except with the two WM variables added. For comparison, two models with Qmci were fitted: one with and one without the two WM variables.ResultThree models achieved cross‐validated classification accuracy of c = 0.90: WM by itself (estimated c = 0.91; Figure 1), WM and a personal narrative task (c = 0.94; Figure 2), and WM and a counting backwards task (c = 0.90). Models with picture description and procedural discourse tasks performed the worst. The QMCI‐only model yielded c = 0.91.ConclusionCombining WM, acoustic and linguistic variables involved in recounting personal memories, and the ability to benefit from prior exposure to a task discriminated CN from CI groups with a high degree of accuracy. This unique combination assesses key cognitive abilities known to be affected early in neurodegenerative diseases, executive functioning and memory, and includes both verbal and nonverbal tasks. Importantly, individuals with varying levels of cognitive functioning, including mild dementia, were able to use the app successfully, and the automated nature of the tool required minimal staff time or expertise.

Structural Correlates of Speech Biomarkers in Alzheimer’s Disease and Mild Cognitive Impairment

AbstractBackgroundIn the present study, we aimed to confirm the association between speech characteristics of individuals with Alzheimer’s disease (AD) and amnestic mild cognitive impairment (MCI) and brain regions, and to investigate the neurobiological basis of speech‐based AD detection.MethodIndividuals with MCI (n = 13, age = 78.3), mild‐to‐moderate AD (n = 12, age = 77.8) underwent a structural MRI scan and engaged in three speech tasks: an interview, sentence repetition, and paragraph recall. From the speech samples, we extracted frequency‐related, intensity‐related, and temporal features which were found to be altered in the course of AD and MCI pathology. We selected nine regions of interest (ROI) which were found to be involved in speech production and reduced in MCI and AD pathology; the Inferior Frontal Gyrus (IFG), the Medial Frontal Gyrus (MFG), the Precuneus, the Inferior Parietal Lobe (IPL), Inferior Temporal Gyrus (ITG), Medial Temporal Gyrus (MTG), the Superior Temporal Gyrus (STG), the Fusiform Gyrus (FFG), and the Cerebellum. The linear mixed‐effect analysis was used to identify the associations between speech features and regional brain volumes.ResultThe Left IPL (χ2 (1) = 7.69, corrected P = 0.033), the Right ITG (χ2 (1) = 11.5, corrected P = 0.0128) the Right FFG (χ2 (1) = 9.35, corrected P = 0.0198) reduced the pitch of voice by about 0.887 ± 0.292, 1.368 ± 0.329, and 1.69 ± 0.429 respectively. The associations between the Right FFG (χ2 (1) = 4.58, corrected P = 0.0900), Left FFG (χ2 (1) = 5.87, corrected P = 0.0675) and loudness were marginally significant, with the reduction of volumes in the Right FFG and Left FFG decreasing the loudness by 0.096 ± 0.041, and 0.1165 ± 0.0454. The marginally significant associations were observed between the Right IPL (χ2 (1) = 4.75, corrected P = 0.0900), the Left FFG (χ2 (1) = 0.39, corrected P = 0.53) and the speech rate, and the atrophy of brain regions decreased the speech rate by 0.0824 ± 0.0363, 0.0535 ± 0.0079.ConclusionThe present study identified the associations between speech biomarkers and regional brain volumes providing clinical validity for the speech‐based Alzheimer’s disease detection.

A multimodal digital biomarker of functional deficits in early‐stage Alzheimer’s disease: results of the RADAR‐AD study

AbstractBackgroundRemote monitoring technologies (RMTs), such as smartphone apps, smartwatches, and in‐home sensors, are rapidly changing the way functional and cognitive performance is measured in Alzheimer’s disease (AD) patients. Here, we present results from the European RADAR‐AD study on the use of multimodal data streams for the identification of functional deficits across all syndromic stages of AD.MethodFour study groups (Healthy controls (HC), preclinical AD (pre. AD), prodromal AD (pro. AD), and mild AD) were included in this cross‐sectional study. The RMT features (gait measures from Timed up and Go (TUG), Dual Task Effect (DTE) using physilog, acoustic features from Speech task in Mezurio, neurocognitive function using Altoida, managing finances with Banking app) with in‐clinic neuropsychological (NP) tests, activities of daily living with Amsterdam IADL and demographics (age, gender, education years) were analyzed for different combinations of the multimodal digital biomarker of disease stage in AD across different pairwise comparisons. The analysis includes data from 175 participants (HC = 67, Pre.AD = 26, Pro.AD = 50, Mild AD = 32) collected for 8 weeks. An extreme gradient boosting (XGBoost) machine learning model was trained to obtain a multimodal biomarker of AD disease stage with repeated cross‐validation (5‐fold with 4 repeats) (Figure 1). This investigation is part of the ongoing RADAR‐AD study.ResultThe multimodal combination of RMTs achieved a mean AUC of &gt; 0.60 in all pairwise comparisons with a mean AUC &gt; 0.65 for HC vs (Pre.AD, Pro.AD and Mild) (Figures 2 and 3). The addition of NP tests increases the performance considerably across all pairwise comparisons except HC vs Pre.AD.ConclusionOur results highlight the advantage of combining RMTs to identify functional deficits in the early stage of AD. In particular, in prodromal and mild AD patients, a combined signal shows much more strength compared to individual tests. This work has received support from the EU/EFPIA Innovative Medicines Initiative Joint Undertaking (grant No 806999). www.imi.europa.eu. This communication reflects the views of the RADAR‐AD consortium and neither IMI nor the European Union and EFPIA are liable for any use that may be made of the information contained herein.

Building a normative sample for smartphone‐based measures of cognition and motor functioning: Study design and initial feasibility findings

AbstractBackgroundSmartphone‐based cognitive and motor tests enable frequent remote assessments in observational research studies and clinical trials, limiting travel burden and increasing access to research for underrepresented groups. One such smartphone tool is the ALLFTD Mobile App, which was designed to capture the clinical features of neurodegenerative diseases via a battery of cognitive, speech, and motor tasks. In this study, we focus on recruiting a demographically diverse normative sample. We have partnered with the UCSF Brain Health Registry (BHR) to recruit 1,000 cognitively healthy participants across the lifespan. Here we describe the study design and initial feasibility results.MethodsWe designed an automated workflow in which participants never interact with the study team except for troubleshooting. BHR participants without reported cognitive impairment were contacted by email to participate in this study (n = 668). Interested participants downloaded the app onto their smartphone and completed in‐app electronic consent. They were asked to complete three ∼30‐minute sessions of cognitive, motor and speech measures over two weeks. Adherence was determined by evaluating the completion rate of app tasks. In a subset who completed a user experience questionnaire (n = 116), responses were quantified using descriptive statistics.ResultsThe current enrollment rate is 23.20% (155/668) of contacted participants. Average age is 62 (SD = 8.2, range = 25‐84). Average adherence is 76%. Participants reported a range of perceived task difficulty: 38% indicated tasks were “very easy” or “somewhat easy”, 26% indicated they were “neutral,” 36% found the tasks “somewhat difficult,” and none reported them to be “very difficult.” The time commitment was reported to be acceptable by 97% of participants, and 20% were willing to complete additional tasks. Over half (59%) of participants indicated that they preferred using a smartphone to complete cognitive assessments over traditional pen‐and‐paper measures.ConclusionsOur preliminary results demonstrate the feasibility and acceptability of recruiting participants, including older adults, through a registry and conducting remote smartphone testing via a fully automated workflow that is highly scalable. Future analysis will study the effects of demographic factors on feasibility and investigate the reliability and validity of the app measures in this sample.

Combining speech biomarkers for cognition and mood improves diagnostic accuracy between MCI and subjective cognitive decliners in a memory clinic sample

AbstractIntroductionDepressive symptoms are a common comorbidity in an old age memory clinic population. Cognitive and affective syndromes coexist, and supposedly modulate each other. While, there have been separate research streams on analysing speech to infer cognitive performance, as well as mood related symptoms, they are seldom investigated side‐by‐side. We present results from a small sub sample of a larger MCI and dementia cohort indicating that modelling both depressive symptoms and cognitive problems from speech can improve the diagnostic accuracy for MCI above speech biomarkers for cognition alone.MethodsThe ki:e Speech Biomarker for cognition (ki:e SB‐C; Tröger et al., 2022) and a free speech task were collected from 35 memory clinic participants (12F) with 10 of them having an MCI diagnosis (compare Table 1). From the free speech tasks (“Tell me about your upcoming vacation plans.”) we extracted speech features using ki:elements proprietary speech analysis engine SIGMA. Free speech features were corrected for cognitive performance, sex and age and then combined into a mood speech biomarker. To benchmark the diagnostic utility of both cognitive and mood speech biomarker side by side, we trained two groups of simple classifiers (tree based classifiers with leave one participant out cross validation) once with SB‐C alone and once with both SB‐C and the newly‐composed mood speech biomarker.ResultsBy combining the speech biomarker for cognition and mood diagnostic performance was significantly improved (AUC of 88%) as compared to classification with the SB‐C alone (AUC of 74%). The mood biomarker correlates significantly with the Geriatric Depression Scale (r = 0.39, p &lt; 0.05). Most important free speech features were jitter, number or pauses, mean pause duration or pronoun rate.ConclusionOverall we present rare evidences that combining speech‐based digital biomarkers tapping both into cognition but also mood‐related symptoms helps to improve diagnostic accuracy in MCI patients. Though just a small feasibility study, the results are clearly encouraging and point towards potential application in both clinical practice as well as clinical research.

Single-pulse electrical stimulation artifact removal using the novel matching pursuit-based artifact reconstruction and removal method (MPARRM)

Objective. Single-pulse electrical stimulation (SPES) has been widely used to probe effective connectivity. However, analysis of the neural response is often confounded by stimulation artifacts. We developed a novel matching pursuit-based artifact reconstruction and removal method (MPARRM) capable of removing artifacts from stimulation-artifact-affected electrophysiological signals. Approach. To validate MPARRM across a wide range of potential stimulation artifact types, we performed a bench-top experiment in which we suspended electrodes in a saline solution to generate 110 types of real-world stimulation artifacts. We then added the generated stimulation artifacts to ground truth signals (stereoelectroencephalography signals from nine human subjects recorded during a receptive speech task), applied MPARRM to the combined signal, and compared the resultant denoised signal with the ground truth signal. We further applied MPARRM to artifact-affected neural signals recorded from the hippocampus while performing SPES on the ipsilateral basolateral amygdala in nine human subjects. Main results. MPARRM could remove stimulation artifacts without introducing spectral leakage or temporal spread. It accommodated variable stimulation parameters and recovered the early response to SPES within a wide range of frequency bands. Specifically, in the early response period (5–10 ms following stimulation onset), we found that the broadband gamma power (70–170 Hz) of the denoised signal was highly correlated with the ground truth signal ( R=0.98±0.02 , Pearson), and the broadband gamma activity of the denoised signal faithfully revealed the responses to the auditory stimuli within the ground truth signal with 94%±1.47% sensitivity and 99%±1.01% specificity. We further found that MPARRM could reveal the expected temporal progression of broadband gamma activity along the anterior-posterior axis of the hippocampus in response to the ipsilateral amygdala stimulation. Significance. MPARRM could faithfully remove SPES artifacts without confounding the electrophysiological signal components, especially during the early-response period. This method can facilitate the understanding of the neural response mechanisms of SPES.