Provider perspectives on equity in use of mobile health autism screening tools.

LAY ABSTRACT: Families may find information about autism online, and health care and education providers may use online tools to screen for autism. However, we do not know if online autism screening tools are easily used by families and providers. We interviewed primary care and educational providers, asking them to review results from online tools that screen for autism. Providers had concerns about how usable and accessible these tools are for diverse families and suggested changes to make tools easier to use.

Consistency and reliability of automated language measures across expressive language samples in autism.

Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder with substantial clinical heterogeneity, especially in language and communication ability. There is a need for validated language outcome measures that show sensitivity to true change for this population. We used Natural Language Processing to analyze expressive language transcripts of 64 highly-verbal children and young adults (age: 6-23 years, mean 12.8 years; 78.1% male) with ASD to examine the validity across language sampling context and test-retest reliability of six previously validated Automated Language Measures (ALMs), including Mean Length of Utterance in Morphemes, Number of Distinct Word Roots, C-units per minute, unintelligible proportion, um rate, and repetition proportion. Three expressive language samples were collected at baseline and again 4 weeks later. These samples comprised interview tasks from the Autism Diagnostic Observation Schedule (ADOS-2) Modules 3 and 4, a conversation task, and a narration task. The influence of language sampling context on each ALM was estimated using either generalized linear mixed-effects models or generalized linear models, adjusted for age, sex, and IQ. The 4 weeks test-retest reliability was evaluated using Lin's Concordance Correlation Coefficient (CCC). The three different sampling contexts were associated with significantly (P < 0.001) different distributions for each ALM. With one exception (repetition proportion), ALMs also showed good test-retest reliability (median CCC: 0.73-0.88) when measured within the same context. Taken in conjunction with our previous work establishing their construct validity, this study demonstrates further critical psychometric properties of ALMs and their promising potential as language outcome measures for ASD research.

"Um" and "Uh" Usage Patterns in Children with Autism: Associations with Measures of Structural and Pragmatic Language Ability.

Pragmatic language difficulties, including unusual filler usage, are common among children with Autism Spectrum Disorder (ASD). This study investigated "um" and "uh" usage in children with ASD and typically developing (TD) controls. We analyzed transcribed Autism Diagnostic Observation Schedule (ADOS) sessions for 182 children (117 ASD, 65 TD), aged 4 to 15. Although the groups did not differ in "uh" usage, the ASD group used fewer "ums" than the TD group. This held true after controlling for age, sex, and IQ. Within ASD, social affect and pragmatic language scores did not predict filler usage; however, structural language scores predicted "um" usage. Lower "um" rates among children with ASD may reflect problems with planning or production rather than pragmatic language.

Mobile and online consumer tools to screen for autism do not promote equity.

LAY ABSTRACT: Many parents wonder if their child might have autism. Many parents use their smartphones to answer health questions. We asked, "How easy or hard is it for parents to use their smartphones to find 'tools' to test their child for signs of autism?" After doing pretend parent searches, we found that only one in 10 search results were tools to test children for autism. These tools were not designed for parents who have low income or other challenges such as low literacy skills, low English proficiency, or not being tech-savvy.

Serving young children with communication disabilities from Latinx backgrounds and their families with equity: Provider perspectives.

INTRODUCTION: Latinx children with communication disorders from birth to age 5 and their families are increasingly served in United States (US) educational and medical settings where longstanding structural barriers threaten their access to equitable assessment and intervention. However, little is known about providers' perceptions serving this highly diverse population as they relate to reducing disparities in care for communication disorders. METHODS: This exploratory qualitative study interviewed 24 speech-language pathologists (SLPs) and early intervention (EI)/early childhood special education (ECSE) developmental specialists serving young Latinx children with communication disorders to offer targeted recommendations toward improving equity. The semi-structured interview included questions regarding communication assessment, diagnostics/eligibility, intervention, interpretation, translation, and solutions to enhance EI/ECSE. Interviews were coded with content analysis using elements of grounded theory, and responses from SLPs in medical versus education settings and from EI/ECSE developmental specialists were compared. Data triangulation was used to validate themes. RESULTS: Analysis revealed the following themes related to provider challenges and resources: family factors, provider factors, cultural and linguistic differences, assessment approaches, eligibility determinations, translation and interpretation, and institutional factors. Few variations in themes between provider types (SLPs vs. EI/ECSE developmental specialists) and settings (medical vs. educational) were found. Providers also offered several policy and practice solutions. CONCLUSIONS: Findings suggest minimal advances in improving equity for young Latinx children with communication disorders over prior decades. Results also indicate that providers may benefit from reflecting on their cultures and biases as well as systemic racism within EI/ECSE.

Pragmatic Language Markers of Autism Diagnosis and Severity.

Purpose: Assessment of pragmatic language difficulties is limited with conventional tests but can be performed with informant reports. We evaluated the performance of a parent-completed language scale in differentiating autism from typical development (TD) and another neurodevelopmental disorder. Specifically, we aimed to gauge the respective values of structural and pragmatic language scores for diagnostic discrimination and for predicting severity of social impairment in autistic children. Method: 174 children aged 7 to 17 (101 with autism, 45 with ADHD, 28 with TD) were evaluated with the ADOS-2 and an abbreviated version of the WISC. Parents completed the Children's Communication Checklist, 2nd Edition (CCC-2) and the Social Responsiveness Scale. CCC-2 mean differences across diagnostic groups were tested with analysis of variance and covariance. Multiple linear regression was used to compare the structural and pragmatic CCC-2 scores in predicting autism symptom severity. Results: Both structural and pragmatic language scores discriminated between the three diagnostic groups, with stronger effects for the pragmatic scores. Pragmatic scores remained robust predictors of ADHD and ASD diagnoses even after accounting for cognitive and structural linguistic differences. Among autistic children, social impairment severity was associated with pragmatic, but not structural, language profiles. Conclusions: In order to characterize pragmatic language, easy to administer parent questionnaires such as the CCC-2 may support clinicians who are considering an autism diagnosis and needing to evaluate and monitor social communication.

A Pseudo-Value Approach to Analyze the Semantic Similarity of the Speech of Children With and Without Autism Spectrum Disorder.

Conversational impairments are well known among people with autism spectrum disorder (ASD), but their measurement requires time-consuming manual annotation of language samples. Natural language processing (NLP) has shown promise in identifying semantic difficulties when compared to clinician-annotated reference transcripts. Our goal was to develop a novel measure of lexico-semantic similarity - based on recent work in natural language processing (NLP) and recent applications of pseudo-value analysis - which could be applied to transcripts of children's conversational language, without recourse to some ground-truth reference document. We hypothesized that: (a) semantic coherence, as measured by this method, would discriminate between children with and without ASD and (b) more variability would be found in the group with ASD. We used data from 70 4- to 8-year-old males with ASD (N = 38) or typically developing (TD; N = 32) enrolled in a language study. Participants were administered a battery of standardized diagnostic tests, including the Autism Diagnostic Observation Schedule (ADOS). ADOS was recorded and transcribed, and we analyzed children's language output during the conversation/interview ADOS tasks. Transcripts were converted to vectors via a word2vec model trained on the Google News Corpus. Pairwise similarity across all subjects and a sample grand mean were calculated. Using a leave-one-out algorithm, a pseudo-value, detailed below, representing each subject's contribution to the grand mean was generated. Means of pseudo-values were compared between the two groups. Analyses were co-varied for nonverbal IQ, mean length of utterance, and number of distinct word roots (NDR). Statistically significant differences were observed in means of pseudo-values between TD and ASD groups (p = 0.007). TD subjects had higher pseudo-value scores suggesting that similarity scores of TD subjects were more similar to the overall group mean. Variance of pseudo-values was greater in the ASD group. Nonverbal IQ, mean length of utterance, or NDR did not account for between group differences. The findings suggest that our pseudo-value-based method can be effectively used to identify specific semantic difficulties that characterize children with ASD without requiring a reference transcript.

Automated Scoring of Tablet-Administered Expressive Language Tests.

Speech and language impairments are common pediatric conditions, with as many as 10% of children experiencing one or both at some point during development. Expressive language disorders in particular often go undiagnosed, underscoring the immediate need for assessments of expressive language that can be administered and scored reliably and objectively. In this paper, we present a set of highly accurate computational models for automatically scoring several common expressive language tasks. In our assessment framework, instructions and stimuli are presented to the child on a tablet computer, which records the child's responses in real time, while a clinician controls the pace and presentation of the tasks using a second tablet. The recorded responses for four distinct expressive language tasks (expressive vocabulary, word structure, recalling sentences, and formulated sentences) are then scored using traditional paper-and-pencil scoring and using machine learning methods relying on a deep neural network-based language representation model. All four tasks can be scored automatically from both clean and verbatim speech transcripts with very high accuracy at the item level (83-99%). In addition, these automated scores correlate strongly and significantly (ρ = 0.76-0.99, p < 0.001) with manual item-level, raw, and scaled scores. These results point to the utility and potential of automated computationally-driven methods of both administering and scoring expressive language tasks for pediatric developmental language evaluation.

Evaluating atypical language in autism using automated language measures.

Measurement of language atypicalities in Autism Spectrum Disorder (ASD) is cumbersome and costly. Better language outcome measures are needed. Using language transcripts, we generated Automated Language Measures (ALMs) and tested their validity. 169 participants (96 ASD, 28 TD, 45 ADHD) ages 7 to 17 were evaluated with the Autism Diagnostic Observation Schedule. Transcripts of one task were analyzed to generate seven ALMs: mean length of utterance in morphemes, number of different word roots (NDWR), um proportion, content maze proportion, unintelligible proportion, c-units per minute, and repetition proportion. With the exception of repetition proportion (p [Formula: see text]), nonparametric ANOVAs showed significant group differences (p[Formula: see text]). The TD and ADHD groups did not differ from each other in post-hoc analyses. With the exception of NDWR, the ASD group showed significantly (p[Formula: see text]) lower scores than both comparison groups. The ALMs were correlated with standardized clinical and language evaluations of ASD. In age- and IQ-adjusted logistic regression analyses, four ALMs significantly predicted ASD status with satisfactory accuracy (67.9-75.5%). When ALMs were combined together, accuracy improved to 82.4%. These ALMs offer a promising approach for generating novel outcome measures.

Automatic Assessment of Language Ability in Children with and without Typical Development.

This study describes a fully automated method of expressive language assessment based on vocal responses of children to a sentence repetition task (SRT), a language test that taps into core language skills. Our proposed method automatically transcribes the vocal responses using a test-specific automatic speech recognition system. From the transcriptions, a regression model predicts the gold standard test scores provided by speech-language pathologists. Our preliminary experimental results on audio recordings of 104 children (43 with typical development and 61 with a neurodevelopmental disorder) verifies the feasibility of the proposed automatic method for predicting gold standard scores on this language test, with averaged mean absolute error of 6.52 (on a observed score range from 0 to 90 with a mean value of 49.56) between observed and predicted ratings.Clinical relevance-We describe the use of fully automatic voice-based scoring in language assessment including the clinical impact this development may have on the field of speech-language pathology. The automated test also creates a technological foundation for the computerization of a broad array of tests for voice-based language assessment.

Automated Scoring of Clinical Expressive Language Evaluation Tasks.

Many clinical assessment instruments used to diagnose language impairments in children include a task in which the subject must formulate a sentence to describe an image using a specific target word. Because producing sentences in this way requires the speaker to integrate syntactic and semantic knowledge in a complex manner, responses are typically evaluated on several different dimensions of appropriateness yielding a single composite score for each response. In this paper, we present a dataset consisting of non-clinically elicited responses for three related sentence formulation tasks, and we propose an approach for automatically evaluating their appropriateness. Using neural machine translation, we generate correct-incorrect sentence pairs to serve as synthetic data in order to increase the amount and diversity of training data for our scoring model. Our scoring model uses transfer learning to facilitate automatic sentence appropriateness evaluation. We further compare custom word embeddings with pre-trained contextualized embeddings serving as features for our scoring model. We find that transfer learning improves scoring accuracy, particularly when using pre-trained contextualized embeddings.

Early Intervention Referral Outcomes for Children at Increased Risk of Experiencing Developmental Delays.

OBJECTIVES: The primary goal was to examine outcomes of Part C early intervention (EI) referrals from a high-risk infant follow-up program and factors associated with success. A secondary aim was to determine how many referred children not evaluated by EI would have likely qualified by either automatically meeting state eligibility criteria with a condition associated with "high-probability" for developmental delays or having test scores evidencing developmental delays. METHODS: Participants included 77 children referred directly to EI from a high-risk infant follow-up program. Scores on the Bayley Scales of Infant and Toddler Development-III, basic demographics, and medical variables were extracted from electronic medical records. Information regarding referral outcomes was gathered via follow-up phone calls to EI programs several months after referral. RESULTS: Results indicate 62% of EI referrals resulted in evaluation, with 69% of those evaluated being found eligible for services. Overall, 34% of referrals resulted in EI enrollment. Of those who were not evaluated, 71% were likely to have qualified based on state eligibility criteria. Follow-up phone call results indicated the majority of families not evaluated (64%) were never successfully contacted by the EI program. CONCLUSIONS: Findings from the present study illustrate the extent of challenges in connecting families with needed EI services and indicate an opportunity for improvement in EI referral processes to increase enrollment for eligible children. Improved communication between referral sources and service providers could support enrollment with detailed documentation of prior testing and explicit reasons for referral. Follow-up calls to confirm receipt of referral may also be necessary.

Modifying a general social-emotional measure for early autism screening.

Objective: Broadband social-emotional screening tools are designed to evaluate a child's social development and interactions. Such tools are expected to have reasonable sensitivity for identifying children at risk for autism spectrum disorder (ASD) but would also likely over-estimate risk for ASD since other conditions can also affect social development. In this study, a subset of ASD items from one general social-emotional screening measure, the Ages & Stages Questionnaires: Social Emotional, 2nd edition, was analyzed to determine if use of an ASD subscale might improve prediction of ASD risk for young children. Methods: The ASD subscale was used with 60 families who had a child referred for an ASD evaluation. Social-emotional screening and ASD screening results were compared with the subsequent results from gold-standard diagnostic testing for ASD at a regional autism center, using contingency matrices. Results: As expected, the social-emotional screening tool identified nearly all of the children in the high-risk clinical sample. Use of the ASD subscale increased specificity for ASD (from 4% to 52%) and demonstrated correct prediction of ASD diagnosis in 70% of ASD cases. Conclusions: These preliminary results suggest that using a subset of ASD-specific items on a social-emotional screening tool can increase the tool's specificity for ASD, by isolating ASD-specific concerns.

Improving ASR Systems for Children with Autism and Language Impairment Using Domain-Focused DNN Transfer Techniques.

This study explores building and improving an automatic speech recognition (ASR) system for children aged 6-9 years and diagnosed with autism spectrum disorder (ASD), language impairment (LI), or both. Working with only 1.5 hours of target data in which children perform the Clinical Evaluation of Language Fundamentals Recalling Sentences task, we apply deep neural network (DNN) weight transfer techniques to adapt a large DNN model trained on the LibriSpeech corpus of adult speech. To begin, we aim to find the best proportional training rates of the DNN layers. Our best configuration yields a 29.38% word error rate (WER). Using this configuration, we explore the effects of quantity and similarity of data augmentation in transfer learning. We augment our training with portions of the OGI Kids' Corpus, adding 4.6 hours of typically developing speakers aged kindergarten through 3rd grade. We find that 2nd grade data alone - approximately the mean age of the target data - outperforms other grades and all the sets combined. Doubling the data for 1st, 2nd, and 3rd grade, we again compare each grade as well as pairs of grades. We find the combination of 1st and 2nd grade performs best at a 26.21% WER.

Part C Early Intervention Enrollment in Low Birth Weight Infants At-Risk for Developmental Delays.

Objectives To investigate enrollment patterns in Part C Early Intervention (EI) for low birth weight (LBW) infants (≤2500 g). A secondary aim is to characterize LBW infants that are not enrolled in EI, but would qualify by meeting criteria for a condition associated with a "high-probability" for developmental delays (i.e., Intraventricular Hemorrhage grade III or higher, Apgar score of ≤5 at 5 min, and/or birth weight of ≤1200 g). Methods Data were gathered from 165 LBW infants participating in a high-risk infant follow-up program. Developmental assessment was completed. Basic demographic information and data regarding enrollment in EI were collected via parent questionnaire. Medical variables were extracted from each infant's electronic medical record. Results 71.5 % of LBW infants were not enrolled in EI. Factors influencing probability of EI enrollment included birth weight, gestational age, developmental test scores, and insurance status. Of the 107 infants living in Oregon who were not enrolled in EI, 42.1 % would qualify for services due to an early medical condition identified in Oregon as a condition associated with a "high-probability" for developmental delays. Conclusions Less than one third of LBW infants were enrolled in EI by their first visit to a high-risk infant follow-up program. Those infants demonstrating developmental delays and public insurance were more likely to be enrolled. The majority of infants who have readily identifiable medical risk factors that qualify them for EI were not enrolled. This study was limited by the constraints implicated by using a clinical sample.

Characteristics of the rhythmic organization of vocal babbling: implications for an amodal linguistic rhythm.

Vocal babbling involves production of rhythmic sequences of a mouth close-open alternation giving the perceptual impression of a sequence of consonant-vowel syllables. Petitto and co-workers have argued vocal babbling rhythm is the same as manual syllabic babbling rhythm, in that it has a frequency of 1 cycle per second. They also assert that adult speech and sign language display the same frequency. However, available evidence suggests that the vocal babbling frequency approximates 3 cycles per second. Both adult spoken language and sign language show higher frequencies than babbling in their respective modalities. No information is currently available on the basic rhythmic parameter of intercyclical variability in either modality. A study of reduplicative babbling by 4 infants and 4 adults producing reduplicated syllables confirms the 3 per second vocal babbling rate, as well as a faster rate in adults, and provides new information on intercyclical variability.