Development of a Machine Learning Algorithm for the Surveillance of Autism Spectrum Disorder

Contributions of Artificial Intelligence to Analysis of Gut Microbiota in Autism Spectrum Disorder: A Systematic Review

BACKGROUND

Autism spectrum disorder (ASD) is a highly heterogeneous neurodevelopmental disorder whose etiology is not known today, but everything indicates that it is multifactorial. For example, genetic and epigenetic factors seem to be involved in the etiology of ASD. In recent years, there has been an increase in studies on the implications of gut microbiota (GM) on the behavior of children with ASD given that dysbiosis in GM may trigger the onset, development and progression of ASD through the microbiota-gut-brain axis. At the same time, significant progress has occurred in the development of artificial intelligence (AI).

METHODS

The aim of the present study was to perform a systematic review of articles using AI to analyze GM in individuals with ASD. In line with the PRISMA model, 12 articles using AI to analyze GM in ASD were selected.

RESULTS

Outcomes reveal that the majority of relevant studies on this topic have been conducted in China (33.3%) and Italy (25%), followed by the Netherlands (16.6%), Mexico (16.6%) and South Korea (8.3%).

CONCLUSIONS

The bacteria Bifidobacterium is the most relevant biomarker with regard to ASD. Although AI provides a very promising approach to data analysis, caution is needed to avoid the over-interpretation of preliminary findings. A first step must be taken to analyze GM in a representative general population and ASD samples in order to obtain a GM standard according to age, sex and country. Thus, more work is required to bridge the gap between AI in mental health research and clinical care in ASD.

Predicting autism traits from baby wellness records: A machine learning approach

LAY ABSTRACT

Timely identification of autism spectrum conditions is a necessity to enable children to receive the most benefit from early interventions. Emerging technological advancements provide avenues for detecting subtle, early indicators of autism from routinely collected health information. This study tested a model that provides a likelihood score for autism diagnosis from baby wellness visit records collected during the first 2 years of life. It included records of 591,989 non-autistic children and 12,846 children with autism. The model identified two-thirds of the autism spectrum condition group (boys 63% and girls 66%). Sex-specific models had several predictive features in common. These included language development, fine motor skills, and social milestones from visits at 12-24 months, mother's age, and lower initial growth but higher last growth measurements. Parental concerns about development or hearing impairment were other predictors. The models differed in other growth measurements and birth parameters. These models can support the detection of early signs of autism in girls and boys by using information routinely recorded during the first 2 years of life.

Transparent deep learning to identify autism spectrum disorders (ASD) in EHR using clinical notes

OBJECTIVE

Machine learning (ML) is increasingly employed to diagnose medical conditions, with algorithms trained to assign a single label using a black-box approach. We created an ML approach using deep learning that generates outcomes that are transparent and in line with clinical, diagnostic rules. We demonstrate our approach for autism spectrum disorders (ASD), a neurodevelopmental condition with increasing prevalence.

METHODS

We use unstructured data from the Centers for Disease Control and Prevention (CDC) surveillance records labeled by a CDC-trained clinician with ASD A1-3 and B1-4 criterion labels per sentence and with ASD cases labels per record using Diagnostic and Statistical Manual of Mental Disorders (DSM5) rules. One rule-based and three deep ML algorithms and six ensembles were compared and evaluated using a test set with 6773 sentences (N = 35 cases) set aside in advance. Criterion and case labeling were evaluated for each ML algorithm and ensemble. Case labeling outcomes were compared also with seven traditional tests.

RESULTS

Performance for criterion labeling was highest for the hybrid BiLSTM ML model. The best case labeling was achieved by an ensemble of two BiLSTM ML models using a majority vote. It achieved 100% precision (or PPV), 83% recall (or sensitivity), 100% specificity, 91% accuracy, and 0.91 F-measure. A comparison with existing diagnostic tests shows that our best ensemble was more accurate overall.

CONCLUSIONS

Transparent ML is achievable even with small datasets. By focusing on intermediate steps, deep ML can provide transparent decisions. By leveraging data redundancies, ML errors at the intermediate level have a low impact on final outcomes.

A Prediction Model of Autism Spectrum Diagnosis from Well-Baby Electronic Data Using Machine Learning

Early detection of autism spectrum disorder (ASD) is crucial for timely intervention, yet diagnosis typically occurs after age three. This study aimed to develop a machine learning model to predict ASD diagnosis using infants' electronic health records obtained through a national screening program and evaluate its accuracy. A retrospective cohort study analyzed health records of 780,610 children, including 1163 with ASD diagnoses. Data encompassed birth parameters, growth metrics, developmental milestones, and familial and post-natal variables from routine wellness visits within the first two years. Using a gradient boosting model with 3-fold cross-validation, 100 parameters predicted ASD diagnosis with an average area under the ROC curve of 0.86 (SD < 0.002). Feature importance was quantified using the Shapley Additive explanation tool. The model identified a high-risk group with a 4.3-fold higher ASD incidence (0.006) compared to the cohort (0.001). Key predictors included failing six milestones in language, social, and fine motor domains during the second year, male gender, parental developmental concerns, non-nursing, older maternal age, lower gestational age, and atypical growth percentiles. Machine learning algorithms capitalizing on preventative care electronic health records can facilitate ASD screening considering complex relations between familial and birth factors, post-natal growth, developmental parameters, and parent concern.

Post-high school outcomes of students with autism spectrum disorder and students with intellectual disability: Utilizing predictive analytics and state data for decision making

This study analyzed the post-high school outcomes of exited high-school students with intellectual disability and autism spectrum disorder from a southwestern U.S. state. A predictive analytics approach was used to analyze these students' post-high school outcomes data, which every state is required to collect each year under U.S. special-education law. Data modeling was conducted with machine learning and logistic regression, which produced two main findings. One, the strongest significant predictors were (a) students spending at least 80% of their instructional days in general education settings and (b) graduating from high school. Two, machine learning models were consistently more accurate in predicting post-high school education or employment than were multilevel logistic regression models. This study concluded with the limitations of the data and predictive-analytic models, and the implications for researchers and state and local education professionals to utilize predictive analytics and state-level post-high school outcomes data for decision making.

An Updated Characterization of Childhood Selective Mutism: Exploring Clinical Features, Treatment Utilization, and School Services

Selective mutism (SM) is a severe but understudied childhood anxiety disorder. Most epidemiological research on SM was conducted decades ago and is limited by small sample sizes. This study analyzes parent-reported clinical data from 230 children with diagnosed and suspected SM to provide current information about the presentation of this disorder. Overall, anxiety and social anxiety symptoms were elevated. Gender ratio, comorbidities and family history of psychopathology were generally aligned with previous research. However, age of onset and diagnosis were both earlier than previously reported, with an average delay of 2 years between onset and diagnosis. The majority of children received therapy and school accommodations for their SM, yet there was large variability in types of interventions. This is the largest survey of children with SM conducted primarily within the US and it constitutes the first systematic inquiry into interventions and accommodations received within clinical and school settings.

A Review of and Roadmap for Data Science and Machine Learning for the Neuropsychiatric Phenotype of Autism

Autism spectrum disorder (autism) is a neurodevelopmental delay that affects at least 1 in 44 children. Like many neurological disorder phenotypes, the diagnostic features are observable, can be tracked over time, and can be managed or even eliminated through proper therapy and treatments. However, there are major bottlenecks in the diagnostic, therapeutic, and longitudinal tracking pipelines for autism and related neurodevelopmental delays, creating an opportunity for novel data science solutions to augment and transform existing workflows and provide increased access to services for affected families. Several efforts previously conducted by a multitude of research labs have spawned great progress toward improved digital diagnostics and digital therapies for children with autism. We review the literature on digital health methods for autism behavior quantification and beneficial therapies using data science. We describe both case-control studies and classification systems for digital phenotyping. We then discuss digital diagnostics and therapeutics that integrate machine learning models of autism-related behaviors, including the factors that must be addressed for translational use. Finally, we describe ongoing challenges and potential opportunities for the field of autism data science. Given the heterogeneous nature of autism and the complexities of the relevant behaviors, this review contains insights that are relevant to neurological behavior analysis and digital psychiatry more broadly.

Performance Characteristics of a Machine-Learning Tool to Predict 7-Day Hospital Readmissions

OBJECTIVES

To develop an institutional machine-learning (ML) tool that utilizes demographic, socioeconomic, and medical information to stratify risk for 7-day readmission after hospital discharge; assess the validity and reliability of the tool; and demonstrate its discriminatory capacity to predict readmissions.

PATIENTS AND METHODS

We performed a combined single-center, cross-sectional, and prospective study of pediatric hospitalists assessing the face and content validity of the developed readmission ML tool. The cross-sectional analyses used data from questionnaire Likert scale responses regarding face and content validity. Prospectively, we compared the discriminatory capacity of provider readmission risk versus the ML tool to predict 7-day readmissions assessed via area under the receiver operating characteristic curve analyses.

RESULTS

Overall, 80% (15 of 20) of hospitalists reported being somewhat to very confident with their ability to accurately predict readmission risk; 53% reported that an ML tool would influence clinical decision-making (face validity). The ML tool variable exhibiting the highest content validity was history of previous 7-day readmission. Prospective provider assessment of risk of 413 discharges showed minimal agreement with the ML tool (κ = 0.104 [95% confidence interval 0.028-0.179]). Both provider gestalt and ML calculations poorly predicted 7-day readmissions (area under the receiver operating characteristic curve: 0.67 vs 0.52; P = .11).

CONCLUSIONS

An ML tool for predicting 7-day hospital readmissions after discharge from the general pediatric ward had limited face and content validity among pediatric hospitalists. Both provider and ML-based determinations of readmission risk were of limited discriminatory value. Before incorporating similar tools into real-time discharge planning, model calibration efforts are needed.

Promising Practices in the Frontiers of Quality Outcome Measurement for Intellectual and Developmental Disability Services

Medicaid Home and Community-Based Services (HCBS) for people with intellectual and developmental disabilities (IDD) are vital for supporting people with IDD to live well in their communities, but there are not set standards for monitoring quality outcomes related to HCBS. In this paper, we propose promising practices for improving the quality of HCBS outcome measurement, based both in the literature and our own experience conducting an extensive U.S. state-level study. Specifically, we discuss: (1) using merged administrative datasets, (2) developing high-quality psychometrics that attend to ecological issues in measurement, (3) using advanced statistical analyses, and (4) creating immersive, user-friendly translational dissemination products. We conclude by suggesting what we see as important new frontiers for researchers to consider in order to enhance the quality of HCBS outcome measurement for people with IDD in the future.

Sensory features in autism: Findings from a large population-based surveillance system

Sensory features (i.e., atypical responses to sensory stimuli) are included in the current diagnostic criteria for autism spectrum disorder. Yet, large population-based studies have not examined these features. This study aimed to determine the prevalence of sensory features among autistic children, and examine associations between sensory features, demographics, and co-occurring problems in other areas. Analysis for this study included a sample comprised of 25,627 four- or eight-year-old autistic children identified through the multistate Autism and Developmental Disabilities Monitoring Network (2006-2014). We calculated the prevalence of sensory features and applied multilevel logistic regression modeling. The majority (74%; 95% confidence interval: 73.5%-74.5%) of the children studied had documented sensory features. In a multivariable model, children who were male and those whose mothers had more years of education had higher odds of documented sensory features. Children from several racial and ethnic minority groups had lower odds of documented sensory features than White, non-Hispanic children. Cognitive problems were not significantly related to sensory features. Problems related to adaptive behavior, emotional states, aggression, attention, fear, motor development, eating, and sleeping were associated with higher odds of having documented sensory features. Results from a large, population-based sample indicate a high prevalence of sensory features in autistic children, as well as relationships between sensory features and co-occurring problems. This study also pointed to potential disparities in the identification of sensory features, which should be examined in future research. Disparities should also be considered clinically to avoid reduced access to supports for sensory features and related functional problems. LAY SUMMARY: In a large, population-based sample of 25,627 autistic children, 74% had documented differences in how they respond to sensation. We also identified significant associations of sensory features with adaptive behavior and problems in other domains. Sensory features were less common among girls, children of color, and children of mothers with fewer years of education, suggesting potential disparities in identification.

Annual Research Review: Translational machine learning for child and adolescent psychiatry

Children and adolescents could benefit from the use of predictive tools that facilitate personalized diagnoses, prognoses, and treatment selection. Such tools have not yet been deployed using traditional statistical methods, potentially due to the limitations of the paradigm and the need to leverage large amounts of digital data. This review will suggest that a machine learning approach could address these challenges and is designed to introduce new readers to the background, methods, and results in the field. A rationale is first introduced followed by an outline of fundamental elements of machine learning approaches. To provide an overview of the use of the techniques in child and adolescent literature, a scoping review of broad trends is then presented. Selected studies are also highlighted in order to draw attention to research areas that are closest to translation and studies that exhibit a high degree of experimental innovation. Limitations to the research, and machine learning approaches generally, are outlined in the penultimate section highlighting issues related to sample sizes, validation, clinical utility, and ethical challenges. Finally, future directions are discussed that could enhance the possibility of clinical implementation and address specific questions relevant to the child and adolescent psychiatry. The review gives a broad overview of the machine learning paradigm in order to highlight the benefits of a shift in perspective towards practically oriented statistical solutions that aim to improve clinical care of children and adolescents.

From high school to postsecondary education, training, and employment: Predicting outcomes for young adults with autism spectrum disorder

BACKGROUND AND AIMS

The fastest growing group of students with disabilities are those with Autism Spectrum Disorder (ASD). States annually report on post-high school outcomes (PSO) of exited students. This study sought to fill two gaps in the literature related to PSO for exited high-school students with ASD and the use of state data and predictive modeling.

METHODS

Data from two states were analyzed using two predictive analytics (PA) methods: multilevel logistic regression and machine learning. The receiver operating characteristic curve (ROC) analysis was used to assess predictive performance.

RESULTS

Data analyses produced two results. One, the strongest predictor of PSO for exited students with ASD was graduating from high school. Two, machine learning performed better than multilevel logistic regression in predicting PSO engagement across the two states.

CONCLUSION

This study contributed two new and important findings to the literature: (a) PA models should be applied to state PSO data because they produce useful information, and (b) PA models are accurate and reliable over time.

IMPLICATIONS

These findings can be used to support state and local educators to make decisions about policies, programs, and practices for exited high school students with ASD, to help them successfully transition to adult life.

Using Machine Learning to Predict Patterns of Employment and Day Program Participation

In this article, we demonstrate the potential of machine learning approaches as inductive analytic tools for expanding our current evidence base for policy making and practice that affects people with intellectual and developmental disabilities (IDD). Using data from the National Core Indicators In-Person Survey (NCI-IPS), a nationally validated annual survey of more than 20,000 nationally representative people with IDD, we fit a series of classification tree and random forest models to predict individuals' employment status and day activity participation as a function of their responses to all other items on the 2017-2018 NCI-IPS. The most accurate model, a random forest classifier, predicted employment outcomes of adults with IDD with an accuracy of 89 percent on the testing sample, and 80 percent on the holdout sample. The most important variable in this prediction was whether or not community employment was a goal in this person's service plan. These results suggest the potential machine learning tools to examine other valued outcomes used in evidence-based policy making to support people with IDD.

Comparison of 2 Case Definitions for Ascertaining the Prevalence of Autism Spectrum Disorder Among 8-Year-Old Children

The Autism and Developmental Disabilities Monitoring (ADDM) Network conducts population-based surveillance of autism spectrum disorder (ASD) among 8-year-old children in multiple US communities. From 2000 to 2016, investigators at ADDM Network sites classified ASD from collected text descriptions of behaviors from medical and educational evaluations which were reviewed and coded by ADDM Network clinicians. It took at least 4 years to publish data from a given surveillance year. In 2018, we developed an alternative case definition utilizing ASD diagnoses or classifications made by community professionals. Using data from surveillance years 2014 and 2016, we compared the new and previous ASD case definitions. Compared with the prevalence based on the previous case definition, the prevalence based on the new case definition was similar for 2014 and slightly lower for 2016. Sex and race/ethnicity prevalence ratios were nearly unchanged. Compared with the previous case definition, the new case definition's sensitivity was 86% and its positive predictive value was 89%. The new case definition does not require clinical review and collects about half as much data, yielding more timely reporting. It also more directly measures community identification of ASD, thus allowing for more valid comparisons among communities, and reduces resource requirements while retaining measurement properties similar to those of the previous definition.

A new classification system for autism based on machine learning of artificial intelligence

BACKGROUND

Autistic Spectrum Disorder (ASD) is a neurodevelopment condition that is normally linked with substantial healthcare costs. Typical ASD screening techniques are time consuming, so the early detection of ASD could reduce such costs and help limit the development of the condition.

OBJECTIVE

We propose an automated approach to detect autistic traits that replaces the scoring function used in current ASD screening with a more intelligent and less subjective approach.

METHODS

The proposed approach employs deep neural networks (DNNs) to detect hidden patterns from previously labelled cases and controls, then applies the knowledge derived to classify the individual being screened. Specificity, sensitivity, and accuracy of the proposed approach are evaluated using ten-fold cross-validation. A comparative analysis has also been conducted to compare the DNNs' performance with other prominent machine learning algorithms.

RESULTS

Results indicate that deep learning technologies can be embedded within existing ASD screening to assist the stakeholders in the early identification of ASD traits.

CONCLUSION

The proposed system will facilitate access to needed support for the social, physical, and educational well-being of the patient and family by making ASD screening more intelligent and accurate.

Is the Combination of ADOS and ADI-R Necessary to Classify ASD? Rethinking the "Gold Standard" in Diagnosing ASD

Diagnosing autism spectrum disorder (ASD) requires extensive clinical expertise and training as well as a focus on differential diagnoses. The diagnostic process is particularly complex given symptom overlap with other mental disorders and high rates of co-occurring physical and mental health concerns. The aim of this study was to conduct a data-driven selection of the most relevant diagnostic information collected from a behavior observation and an anamnestic interview in two clinical samples of children/younger adolescents and adolescents/adults with suspected ASD. Via random forests, the present study discovered patterns of symptoms in the diagnostic data of 2310 participants (46% ASD, 54% non-ASD, age range 4-72 years) using data from the combined Autism Diagnostic Observation Schedule (ADOS) and Autism Diagnostic Interview-Revised (ADI-R) and ADOS data alone. Classifiers built on reduced subsets of diagnostic features yield satisfactory sensitivity and specificity values. For adolescents/adults specificity values were lower compared to those for children/younger adolescents. The models including ADOS and ADI-R data were mainly built on ADOS items and in the adolescent/adult sample the classifier including only ADOS items performed even better than the classifier including information from both instruments. Results suggest that reduced subsets of ADOS and ADI-R items may suffice to effectively differentiate ASD from other mental disorders. The imbalance of ADOS and ADI-R items included in the models leads to the assumption that, particularly in adolescents and adults, the ADI-R may play a lesser role than current behavior observations.

Informing Developmental Milestone Achievement for Children With Autism: Machine Learning Approach

Background

Care for children with autism spectrum disorder (ASD) can be challenging for families and medical care systems. This is especially true in low- and- middle-income countries such as Bangladesh. To improve family–practitioner communication and developmental monitoring of children with ASD, mCARE (Mobile-Based Care for Children with Autism Spectrum Disorder Using Remote Experience Sampling Method) was developed. Within this study, mCARE was used to track child milestone achievement and family sociodemographic assets to inform mCARE feasibility/scalability and family asset–informed practitioner recommendations.

Objective

The objectives of this paper are threefold. First, it documents how mCARE can be used to monitor child milestone achievement. Second, it demonstrates how advanced machine learning models can inform our understanding of milestone achievement in children with ASD. Third, it describes family/child sociodemographic factors that are associated with earlier milestone achievement in children with ASD (across 5 machine learning models).

Methods

Using mCARE-collected data, this study assessed milestone achievement in 300 children with ASD from Bangladesh. In this study, we used 4 supervised machine learning algorithms (decision tree, logistic regression, K-nearest neighbor [KNN], and artificial neural network [ANN]) and 1 unsupervised machine learning algorithm (K-means clustering) to build models of milestone achievement based on family/child sociodemographic details. For analyses, the sample was randomly divided in half to train the machine learning models and then their accuracy was estimated based on the other half of the sample. Each model was specified for the following milestones: Brushes teeth, Asks to use the toilet, Urinates in the toilet or potty, and Buttons large buttons.

Results

This study aimed to find a suitable machine learning algorithm for milestone prediction/achievement for children with ASD using family/child sociodemographic characteristics. For Brushes teeth, the 3 supervised machine learning models met or exceeded an accuracy of 95% with logistic regression, KNN, and ANN as the most robust sociodemographic predictors. For Asks to use toilet, 84.00% accuracy was achieved with the KNN and ANN models. For these models, the family sociodemographic predictors of “family expenditure” and “parents’ age” accounted for most of the model variability. The last 2 parameters, Urinates in toilet or potty and Buttons large buttons, had an accuracy of 91.00% and 76.00%, respectively, in ANN. Overall, the ANN had a higher accuracy (above ~80% on average) among the other algorithms for all the parameters. Across the models and milestones, “family expenditure,” “family size/type,” “living places,” and “parent’s age and occupation” were the most influential family/child sociodemographic factors.

Conclusions

mCARE was successfully deployed in a low- and middle-income country (ie, Bangladesh), providing parents and care practitioners a mechanism to share detailed information on child milestones achievement. Using advanced modeling techniques this study demonstrates how family/child sociodemographic elements can inform child milestone achievement. Specifically, families with fewer sociodemographic resources reported later milestone attainment. Developmental science theories highlight how family/systems can directly influence child development and this study provides a clear link between family resources and child developmental progress. Clinical implications for this work could include supporting the larger family system to improve child milestone achievement.

Virulence factor-related gut microbiota genes and immunoglobulin A levels as novel markers for machine learning-based classification of autism spectrum disorder

Autism spectrum disorder (ASD) is a neurodevelopmental condition for which early identification and intervention is crucial for optimum prognosis. Our previous work showed gut Immunoglobulin A (IgA) to be significantly elevated in the gut lumen of children with ASD compared to typically developing (TD) children. Gut microbiota variations have been reported in ASD, yet not much is known about virulence factor-related gut microbiota (VFGM) genes. Upon determining the VFGM genes distinguishing ASD from TD, this study is the first to utilize VFGM genes and IgA levels for a machine learning-based classification of ASD. Sequence comparisons were performed of metagenome datasets from children with ASD (n = 43) and TD children (n = 31) against genes in the virulence factor database. VFGM gene composition was associated with ASD phenotype. VFGM gene diversity was higher in children with ASD and positively correlated with IgA content. As Group B streptococcus (GBS) genes account for the highest proportion of 24 different VFGMs between ASD and TD and positively correlate with gut IgA, GBS genes were used in combination with IgA and VFGMs diversity to distinguish ASD from TD. Given that VFGM diversity, increases in IgA, and ASD-enriched VFGM genes were independent of sex and gastrointestinal symptoms, a classification method utilizing them will not pertain only to a specific subgroup of ASD. By introducing the classification value of VFGM genes and considering that VFs can be isolated in pregnant women and newborns, these findings provide a novel machine learning-based early risk identification method for ASD.

Autism spectrum disorder/Takiwātanga: An Integrated Data Infrastructure-based approach to autism spectrum disorder research in New Zealand

New Zealand has few estimates of the prevalence of autism spectrum disorder and no national registry. The use of administrative data sources is expanding and could be useful in autism spectrum disorder research. However, the extent to which autism spectrum disorder can be captured in these data sources is unknown. In this study, we utilised three linked administrative health data sources from the Integrated Data Infrastructure to identify cases of autism spectrum disorder among New Zealand children and young people. We then investigated the extent to which a range of mental health, neurodevelopmental and related problems co-occur with autism spectrum disorder. In total, 9555 unique individuals aged 0–24 with autism spectrum disorder were identified. The identification rate for 8-year-olds was 1 in 102. Co-occurring mental health or related problems were noted in 68% of the autism spectrum disorder group. The most common co-occurring conditions were intellectual disability, disruptive behaviours and emotional problems. Although data from the Integrated Data Infrastructure may currently undercount cases of autism spectrum disorder, they could be useful for monitoring service and treatment-related trends, types of co-occurring conditions and for examining social outcomes. With further refinement, the Integrated Data Infrastructure could prove valuable for informing the national incidence and prevalence of autism spectrum disorder and the long-term effectiveness of clinical guidelines and interventions for this group.

A predictive model for paediatric autism screening

Autism spectrum disorder is an umbrella term for a group of neurodevelopmental disorders that is associated with impairments to social interaction, communication, and behaviour. Typically, autism spectrum disorder is first detected with a screening tool (e.g. modified checklist for autism in toddlers). However, the interpretation of autism spectrum disorder behavioural symptoms varies across cultures: the sensitivity of modified checklist for autism in toddlers is as low as 25 per cent in Sri Lanka. A culturally sensitive screening tool called pictorial autism assessment schedule has overcome this problem. Low- and middle-income countries have a shortage of mental health specialists, which is a key barrier for obtaining an early autism spectrum disorder diagnosis. Early identification of autism spectrum disorder enables intervention before atypical patterns of behaviour and brain function become established. This article proposes a culturally sensitive autism spectrum disorder screening mobile application. The proposed application embeds an intelligent machine learning model and uses a clinically validated symptom checklist to monitor and detect autism spectrum disorder in low- and middle-income countries for the first time. Machine learning models were trained on clinical pictorial autism assessment schedule data and their predictive performance was evaluated, which demonstrated that the random forest was the optimal classifier (area under the receiver operating characteristic (0.98)) for embedding into the mobile screening tool. In addition, feature selection demonstrated that many pictorial autism assessment schedule questions are redundant and can be removed to optimise the screening process.

Identifying predictive features of autism spectrum disorders in a clinical sample of adolescents and adults using machine learning

Diagnosing autism spectrum disorders (ASD) is a complicated, time-consuming process which is particularly challenging in older individuals. One of the most widely used behavioral diagnostic tools is the Autism Diagnostic Observation Schedule (ADOS). Previous work using machine learning techniques suggested that ASD detection in children can be achieved with substantially fewer items than the original ADOS. Here, we expand on this work with a specific focus on adolescents and adults as assessed with the ADOS Module 4. We used a machine learning algorithm (support vector machine) to examine whether ASD detection can be improved by identifying a subset of behavioral features from the ADOS Module 4 in a routine clinical sample of N = 673 high-functioning adolescents and adults with ASD (n = 385) and individuals with suspected ASD but other best-estimate or no psychiatric diagnoses (n = 288). We identified reduced subsets of 5 behavioral features for the whole sample as well as age subgroups (adolescents vs. adults) that showed good specificity and sensitivity and reached performance close to that of the existing ADOS algorithm and the full ADOS, with no significant differences in overall performance. These results may help to improve the complicated diagnostic process of ASD by encouraging future efforts to develop novel diagnostic instruments for ASD detection based on the identified constructs as well as aiding clinicians in the difficult question of differential diagnosis.

New technologies and future trends

Numerous technologies have been introduced for the diagnosis, treatment, and management of patients with neurologic disorders, offering the promise of early diagnosis, tailored and individualized interventions, improvement in quality of life, and restoration of neurologic function. Many of these technologies have become available commercially without having been evaluated by rigorous clinical trials and regulatory reviews, or at the least by peer review of results submitted for publication. A subset is intended to assess, assist, and monitor cognitive functions, motor skills, and autonomic functions and as such may be applicable to persons with developmental disabilities. Barriers that have previously limited the use of technologies by persons with neurodevelopmental disabilities are disappearing as new technologies that have the potential to substantially augment diagnosis and interventions to enhance the daily lives of persons with these disorders are emerging. While recent and future advances in technology have the potential to transform their lives, cautious and thoughtful evaluation is needed to ensure the technologies provide maximal value. As such, further work is needed to demonstrate feasibility, efficacy, and cost-effectiveness, and technologies should be designed to be optimized for individual use.

Machine Learning to Study Social Interaction Difficulties in ASD

Autism Spectrum Disorder (ASD) is a spectrum of neurodevelopmental conditions characterized by difficulties in social communication and social interaction as well as repetitive behaviors and restricted interests. Prevalence rates have been rising, and existing diagnostic methods are both extremely time and labor consuming. There is an urgent need for more economic and objective automatized diagnostic tools that are independent of language and experience of the diagnostician and that can help deal with the complexity of the autistic phenotype. Technological advancements in machine learning are offering a potential solution, and several studies have employed computational approaches to classify ASD based on phenomenological, behavioral or neuroimaging data. Despite of being at the core of ASD diagnosis and having the potential to be used as a behavioral marker for machine learning algorithms, only recently have movement parameters been used as features in machine learning classification approaches. In a proof-of-principle analysis of data from a social interaction study we trained a classification algorithm on intrapersonal synchrony as an automatically and objectively measured phenotypic feature from 29 autistic and 29 typically developed individuals to differentiate those individuals with ASD from those without ASD. Parameters included nonverbal motion energy values from 116 videos of social interactions. As opposed to previous studies to date, our classification approach has been applied to non-verbal behavior objectively captured during naturalistic and complex interactions with a real human interaction partner assuring high external validity. A machine learning approach lends itself particularly for capturing heterogeneous and complex behavior in real social interactions and will be essential in developing automatized and objective classification methods in ASD.

A comparison of machine learning algorithms for the surveillance of autism spectrum disorder

Objective

The Centers for Disease Control and Prevention (CDC) coordinates a labor-intensive process to measure the prevalence of autism spectrum disorder (ASD) among children in the United States. Random forests methods have shown promise in speeding up this process, but they lag behind human classification accuracy by about 5%. We explore whether more recently available document classification algorithms can close this gap.

Materials and methods

Using data gathered from a single surveillance site, we applied 8 supervised learning algorithms to predict whether children meet the case definition for ASD based solely on the words in their evaluations. We compared the algorithms’ performance across 10 random train-test splits of the data, using classification accuracy, F₁ score, and number of positive calls to evaluate their potential use for surveillance.

Results

Across the 10 train-test cycles, the random forest and support vector machine with Naive Bayes features (NB-SVM) each achieved slightly more than 87% mean accuracy. The NB-SVM produced significantly more false negatives than false positives (P = 0.027), but the random forest did not, making its prevalence estimates very close to the true prevalence in the data. The best-performing neural network performed similarly to the random forest on both measures.

Discussion

The random forest performed as well as more recently available models like the NB-SVM and the neural network, and it also produced good prevalence estimates. NB-SVM may not be a good candidate for use in a fully-automated surveillance workflow due to increased false negatives. More sophisticated algorithms, like hierarchical convolutional neural networks, may not be feasible to train due to characteristics of the data. Current algorithms might perform better if the data are abstracted and processed differently and if they take into account information about the children in addition to their evaluations.

Conclusion

Deep learning models performed similarly to traditional machine learning methods at predicting the clinician-assigned case status for CDC’s autism surveillance system. While deep learning methods had limited benefit in this task, they may have applications in other surveillance systems.

Review of Machine Learning Algorithms for Diagnosing Mental Illness

OBJECTIVE

Enhanced technology in computer and internet has driven scale and quality of data to be improved in various areas including healthcare sectors. Machine Learning (ML) has played a pivotal role in efficiently analyzing those big data, but a general misunderstanding of ML algorithms still exists in applying them (e.g., ML techniques can settle a problem of small sample size, or deep learning is the ML algorithm). This paper reviewed the research of diagnosing mental illness using ML algorithm and suggests how ML techniques can be employed and worked in practice.

METHODS

Researches about mental illness diagnostic using ML techniques were carefully reviewed. Five traditional ML algorithms-Support Vector Machines (SVM), Gradient Boosting Machine (GBM), Random Forest, Naïve Bayes, and K-Nearest Neighborhood (KNN)-frequently used for mental health area researches were systematically organized and summarized.

RESULTS

Based on literature review, it turned out that Support Vector Machines (SVM), Gradient Boosting Machine (GBM), Random Forest, Naïve Bayes, and K-Nearest Neighborhood (KNN) were frequently employed in mental health area, but many researchers did not clarify the reason for using their ML algorithm though every ML algorithm has its own advantages. In addition, there were several studies to apply ML algorithms without fully understanding the data characteristics.

CONCLUSION

Researchers using ML algorithms should be aware of the properties of their ML algorithms and the limitation of the results they obtained under restricted data conditions. This paper provides useful information of the properties and limitation of each ML algorithm in the practice of mental health.

A new machine learning model based on induction of rules for autism detection

Autism spectrum disorder is a developmental disorder that describes certain challenges associated with communication (verbal and non-verbal), social skills, and repetitive behaviors. Typically, autism spectrum disorder is diagnosed in a clinical environment by licensed specialists using procedures which can be lengthy and cost-ineffective. Therefore, scholars in the medical, psychology, and applied behavioral science fields have in recent decades developed screening methods such as the Autism Spectrum Quotient and Modified Checklist for Autism in Toddlers for diagnosing autism and other pervasive developmental disorders. The accuracy and efficiency of these screening methods rely primarily on the experience and knowledge of the user, as well as the items designed in the screening method. One promising direction to improve the accuracy and efficiency of autism spectrum disorder detection is to build classification systems using intelligent technologies such as machine learning. Machine learning offers advanced techniques that construct automated classifiers that can be exploited by users and clinicians to significantly improve sensitivity, specificity, accuracy, and efficiency in diagnostic discovery. This article proposes a new machine learning method called Rules-Machine Learning that not only detects autistic traits of cases and controls but also offers users knowledge bases (rules) that can be utilized by domain experts in understanding the reasons behind the classification. Empirical results on three data sets related to children, adolescents, and adults show that Rules-Machine Learning offers classifiers with higher predictive accuracy, sensitivity, harmonic mean, and specificity than those of other machine learning approaches such as Boosting, Bagging, decision trees, and rule induction.

Period of Measurement in Time-Series Predictions of Disease Counts from 2007 to 2017 in Northern Nevada: Analytics Experiment

BACKGROUND

The literature in statistics presents methods by which autocorrelation can identify the best period of measurement to improve the performance of a time-series prediction. The period of measurement plays an important role in improving the performance of disease-count predictions. However, from the operational perspective in public health surveillance, there is a limitation to the length of the measurement period that can offer meaningful and valuable predictions.

OBJECTIVE

This study aimed to establish a method that identifies the shortest period of measurement without significantly decreasing the prediction performance for time-series analysis of disease counts.

METHODS

The data used in this evaluation include disease counts from 2007 to 2017 in northern Nevada. The disease counts for chlamydia, salmonella, respiratory syncytial virus, gonorrhea, viral meningitis, and influenza A were predicted.

RESULTS

Our results showed that autocorrelation could not guarantee the best performance for prediction of disease counts. However, the proposed method with the change-point analysis suggests a period of measurement that is operationally acceptable and performance that is not significantly different from the best prediction.

CONCLUSIONS

The use of change-point analysis with autocorrelation provides the best and most practical period of measurement.

Automated Extraction of Diagnostic Criteria From Electronic Health Records for Autism Spectrum Disorders: Development, Evaluation, and Application

BACKGROUND

Electronic health records (EHRs) bring many opportunities for information utilization. One such use is the surveillance conducted by the Centers for Disease Control and Prevention to track cases of autism spectrum disorder (ASD). This process currently comprises manual collection and review of EHRs of 4- and 8-year old children in 11 US states for the presence of ASD criteria. The work is time-consuming and expensive.

OBJECTIVE

Our objective was to automatically extract from EHRs the description of behaviors noted by the clinicians in evidence of the diagnostic criteria in the Diagnostic and Statistical Manual of Mental Disorders (DSM). Previously, we reported on the classification of entire EHRs as ASD or not. In this work, we focus on the extraction of individual expressions of the different ASD criteria in the text. We intend to facilitate large-scale surveillance efforts for ASD and support analysis of changes over time as well as enable integration with other relevant data.

METHODS

We developed a natural language processing (NLP) parser to extract expressions of 12 DSM criteria using 104 patterns and 92 lexicons (1787 terms). The parser is rule-based to enable precise extraction of the entities from the text. The entities themselves are encompassed in the EHRs as very diverse expressions of the diagnostic criteria written by different people at different times (clinicians, speech pathologists, among others). Due to the sparsity of the data, a rule-based approach is best suited until larger datasets can be generated for machine learning algorithms.

RESULTS

We evaluated our rule-based parser and compared it with a machine learning baseline (decision tree). Using a test set of 6636 sentences (50 EHRs), we found that our parser achieved 76% precision, 43% recall (ie, sensitivity), and >99% specificity for criterion extraction. The performance was better for the rule-based approach than for the machine learning baseline (60% precision and 30% recall). For some individual criteria, precision was as high as 97% and recall 57%. Since precision was very high, we were assured that criteria were rarely assigned incorrectly, and our numbers presented a lower bound of their presence in EHRs. We then conducted a case study and parsed 4480 new EHRs covering 10 years of surveillance records from the Arizona Developmental Disabilities Surveillance Program. The social criteria (A1 criteria) showed the biggest change over the years. The communication criteria (A2 criteria) did not distinguish the ASD from the non-ASD records. Among behaviors and interests criteria (A3 criteria), 1 (A3b) was present with much greater frequency in the ASD than in the non-ASD EHRs.

CONCLUSIONS

Our results demonstrate that NLP can support large-scale analysis useful for ASD surveillance and research. In the future, we intend to facilitate detailed analysis and integration of national datasets.

GapMap: Enabling Comprehensive Autism Resource Epidemiology

BACKGROUND

For individuals with autism spectrum disorder (ASD), finding resources can be a lengthy and difficult process. The difficulty in obtaining global, fine-grained autism epidemiological data hinders researchers from quickly and efficiently studying large-scale correlations among ASD, environmental factors, and geographical and cultural factors.

OBJECTIVE

The objective of this study was to define resource load and resource availability for families affected by autism and subsequently create a platform to enable a more accurate representation of prevalence rates and resource epidemiology.

METHODS

We created a mobile application, GapMap, to collect locational, diagnostic, and resource use information from individuals with autism to compute accurate prevalence rates and better understand autism resource epidemiology. GapMap is hosted on AWS S3, running on a React and Redux front-end framework. The backend framework is comprised of an AWS API Gateway and Lambda Function setup, with secure and scalable end points for retrieving prevalence and resource data, and for submitting participant data. Measures of autism resource scarcity, including resource load, resource availability, and resource gaps were defined and preliminarily computed using simulated or scraped data.

RESULTS

The average distance from an individual in the United States to the nearest diagnostic center is approximately 182 km (50 miles), with a standard deviation of 235 km (146 miles). The average distance from an individual with ASD to the nearest diagnostic center, however, is only 32 km (20 miles), suggesting that individuals who live closer to diagnostic services are more likely to be diagnosed.

CONCLUSIONS

This study confirmed that individuals closer to diagnostic services are more likely to be diagnosed and proposes GapMap, a means to measure and enable the alleviation of increasingly overburdened diagnostic centers and resource-poor areas where parents are unable to diagnose their children as quickly and easily as needed. GapMap will collect information that will provide more accurate data for computing resource loads and availability, uncovering the impact of resource epidemiology on age and likelihood of diagnosis, and gathering localized autism prevalence rates.