Mobilized telemedicine is becoming a key, and even necessary, facet of both precision health and precision medicine. In this study, we evaluate the capability and potential of a crowd of virtual workers—defined as vetted members of popular crowdsourcing platforms—to aid in the task of diagnosing autism. We evaluate workers when crowdsourcing the task of providing categorical ordinal behavioral ratings to unstructured public YouTube videos of children with autism and neurotypical controls. To evaluate emerging patterns that are consistent across independent crowds, we target workers from distinct geographic loci on two crowdsourcing platforms: an international group of workers on Amazon Mechanical Turk (MTurk) (N = 15) and Microworkers from Bangladesh (N = 56), Kenya (N = 23), and the Philippines (N = 25). We feed worker responses as input to a validated diagnostic machine learning classifier trained on clinician-filled electronic health records. We find that regardless of crowd platform or targeted country, workers vary in the average confidence of the correct diagnosis predicted by the classifier. The best worker responses produce a mean probability of the correct class above 80% and over one standard deviation above 50%, accuracy and variability on par with experts according to prior studies. There is a weak correlation between mean time spent on task and mean performance (r = 0.358, p = 0.005). These results demonstrate that while the crowd can produce accurate diagnoses, there are intrinsic differences in crowdworker ability to rate behavioral features. We propose a novel strategy for recruitment of crowdsourced workers to ensure high quality diagnostic evaluations of autism, and potentially many other pediatric behavioral health conditions. Our approach represents a viable step in the direction of crowd-based approaches for more scalable and affordable precision medicine.

Activity recognition computer vision algorithms can be used to detect the presence of autism-related behaviors, including what are termed “restricted and repetitive behaviors”, or stimming, by diagnostic instruments. Examples of stimming include hand flapping, spinning, and head banging. One of the most significant bottlenecks for implementing such classifiers is the lack of sufficiently large training sets of human behavior specific to pediatric developmental delays. The data that do exist are usually recorded with a handheld camera which is itself shaky or even moving, posing a challenge for traditional feature representation approaches for activity detection which capture the camera's motion as a feature. To address these issues, we first document the advantages and limitations of current feature representation techniques for activity recognition when applied to head banging detection. We then propose a feature representation consisting exclusively of head pose keypoints. We create a computer vision classifier for detecting head banging in home videos using a time-distributed convolutional neural network (CNN) in which a single CNN extracts features from each frame in the input sequence, and these extracted features are fed as input to a long short-term memory (LSTM) network. On the binary task of predicting head banging and no head banging within videos from the Self Stimulatory Behaviour Dataset (SSBD), we reach a mean F1-score of 90.77% using 3-fold cross validation (with individual fold F1-scores of 83.3%, 89.0%, and 100.0%) when ensuring that no child who appeared in the train set was in the test set for all folds. This work documents a successful process for training a computer vision classifier which can detect a particular human motion pattern with few training examples and even when the camera recording the source clip is unstable. The process of engineering useful feature representations by visually inspecting the representations, as described here, can be a useful practice by designers and developers of interactive systems detecting human motion patterns for use in mobile and ubiquitous interactive systems.

Background: Autism spectrum disorder (ASD) is a widespread neurodevelopmental condition with a range of potential causes and symptoms. Standard diagnostic mechanisms for ASD, which involve lengthy parent questionnaires and clinical observation, often result in long waiting times for results. Recent advances in computer vision and mobile technology hold potential for speeding up the diagnostic process by enabling computational analysis of behavioral and social impairments from home videos. Such techniques can improve objectivity and contribute quantitatively to the diagnostic process. Objective: In this work, we evaluate whether home videos collected from a game-based mobile app can be used to provide diagnostic insights into ASD. To the best of our knowledge, this is the first study attempting to identify potential social indicators of ASD from mobile phone videos without the use of eye-tracking hardware, manual annotations, and structured scenarios or clinical environments. Methods: Here, we used a mobile health app to collect over 11 hours of video footage depicting 95 children engaged in gameplay in a natural home environment. We used automated data set annotations to analyze two social indicators that have previously been shown to differ between children with ASD and their neurotypical (NT) peers: (1) gaze fixation patterns, which represent regions of an individual’s visual focus and (2) visual scanning methods, which refer to the ways in which individuals scan their surrounding environment. We compared the gaze fixation and visual scanning methods used by children during a 90-second gameplay video to identify statistically significant differences between the 2 cohorts; we then trained a long short-term memory (LSTM) neural network to determine if gaze indicators could be predictive of ASD. Results: Our results show that gaze fixation patterns differ between the 2 cohorts; specifically, we could identify 1 statistically significant region of fixation (P<.001). In addition, we also demonstrate that there are unique visual scanning patterns that exist for individuals with ASD when compared to NT children (P<.001). A deep learning model trained on coarse gaze fixation annotations demonstrates mild predictive power in identifying ASD. Conclusions: Ultimately, our study demonstrates that heterogeneous video data sets collected from mobile devices hold potential for quantifying visual patterns and providing insights into ASD. We show the importance of automated labeling techniques in generating large-scale data sets while simultaneously preserving the privacy of participants, and we demonstrate that specific social engagement indicators associated with ASD can be identified and characterized using such data.

Standard medical diagnosis of mental health conditions requires licensed experts who are increasingly outnumbered by those at risk, limiting reach. We test the hypothesis that a trustworthy crowd of non-experts can efficiently annotate behavioral features needed for accurate machine learning detection of the common childhood developmental disorder Autism Spectrum Disorder (ASD) for children under 8 years old. We implement a novel process for identifying and certifying a trustworthy distributed workforce for video feature extraction, selecting a workforce of 102 workers from a pool of 1,107. Two previously validated ASD logistic regression classifiers, evaluated against parent-reported diagnoses, were used to assess the accuracy of the trusted crowd’s ratings of unstructured home videos. A representative balanced sample (N = 50 videos) of videos were evaluated with and without face box and pitch shift privacy alterations, with AUROC and AUPRC scores > 0.98. With both privacy-preserving modifications, sensitivity is preserved (96.0%) while maintaining specificity (80.0%) and accuracy (88.0%) at levels comparable to prior classification methods without alterations. We find that machine learning classification from features extracted by a certified nonexpert crowd achieves high performance for ASD detection from natural home videos of the child at risk and maintains high sensitivity when privacy-preserving mechanisms are applied. These results suggest that privacy-safeguarded crowdsourced analysis of short home videos can help enable rapid and mobile machine-learning detection of developmental delays in children.

Background: Automated emotion classification could aid those who struggle to recognize emotions, including children with developmental behavioral conditions such as autism. However, most computer vision emotion recognition models are trained on adult emotion and therefore underperform when applied to child faces. Objective: We designed a strategy to gamify the collection and labeling of child emotion–enriched images to boost the performance of automatic child emotion recognition models to a level closer to what will be needed for digital health care approaches. Methods: We leveraged our prototype therapeutic smartphone game, GuessWhat, which was designed in large part for children with developmental and behavioral conditions, to gamify the secure collection of video data of children expressing a variety of emotions prompted by the game. Independently, we created a secure web interface to gamify the human labeling effort, called HollywoodSquares, tailored for use by any qualified labeler. We gathered and labeled 2155 videos, 39,968 emotion frames, and 106,001 labels on all images. With this drastically expanded pediatric emotion–centric database (>30 times larger than existing public pediatric emotion data sets), we trained a convolutional neural network (CNN) computer vision classifier of happy, sad, surprised, fearful, angry, disgust, and neutral expressions evoked by children. Results: The classifier achieved a 66.9% balanced accuracy and 67.4% F1-score on the entirety of the Child Affective Facial Expression (CAFE) as well as a 79.1% balanced accuracy and 78% F1-score on CAFE Subset A, a subset containing at least 60% human agreement on emotions labels. This performance is at least 10% higher than all previously developed classifiers evaluated against CAFE, the best of which reached a 56% balanced accuracy even when combining “anger” and “disgust” into a single class. Conclusions: This work validates that mobile games designed for pediatric therapies can generate high volumes of domain-relevant data sets to train state-of-the-art classifiers to perform tasks helpful to precision health efforts.

Autism Spectrum Disorder is a neuropsychiatric condition affecting 53 million children worldwide and for which early diagnosis is critical to the outcome of behavior therapies. Machine learning applied to features manually extracted from readily accessible videos (e.g., from smartphones) has the potential to scale this diagnostic process. However, nearly unavoidable variability in video quality can lead to missing features that degrade algorithm performance. To manage this uncertainty, we evaluated the impact of missing values and feature imputation methods on two previously published autism detection classifiers, trained on standard-of-care instrument scoresheets and tested on ratings of 140 children videos from YouTube. We compare the baseline method of listwise deletion to classic univariate and multivariate techniques. We also introduce a feature replacement method that, based on a score, selects a feature from an expanded dataset to fill-in the missing value. The replacement feature selected can be identical for all records (general) or automatically adjusted to the record considered (dynamic). Our results show that general and dynamic feature replacement methods achieve a higher performance than classic univariate and multivariate methods, supporting the hypothesis that algorithmic management can maintain the fidelity of video-based diagnostics in the face of missing values and variable video quality.