Active Viewing Facilitates Gaze to the Eye Region in Young Children with Autism Spectrum Disorder

A novel multi-modal model to assist the diagnosis of autism spectrum disorder using eye-tracking data

Background and objective

Timely and accurate detection of Autism Spectrum Disorder (ASD) is essential for early intervention and improved patient outcomes. This study aims to harness the power of machine learning (ML) techniques to improve ASD detection by incorporating temporal eye-tracking data. We developed a novel ML model to leverage eye scan paths, sequences of distances of eye movement, and a sequence of fixation durations, enhancing the temporal aspect of the analysis for more effective ASD identification.

Methods

We utilized a dataset of eye-tracking data without augmentation to train our ML model, which consists of a CNN-GRU-ANN architecture. The model was trained using gaze maps, the sequences of distances between eye fixations, and durations of fixations and saccades. Additionally, we employed a validation dataset to assess the model's performance and compare it with other works.

Results

Our ML model demonstrated superior performance in ASD detection compared to the VGG-16 model. By incorporating temporal information from eye-tracking data, our model achieved higher accuracy, precision, and recall. The novel addition of sequence-based features allowed our model to effectively distinguish between ASD and typically developing individuals, achieving an impressive precision value of 93.10% on the validation dataset.

Conclusion

This study presents an ML-based approach to ASD detection by utilizing machine learning techniques and incorporating temporal eye-tracking data. Our findings highlight the potential of temporal analysis for improved ASD detection and provide a promising direction for further advancements in the field of eye-tracking-based diagnosis and intervention for neurodevelopmental disorders.

Gaze Following as an Early Diagnostic Marker of Autism in a New Word Learning Task in Toddlers

The aim was to test the use of eye-tracking methodology for the early detection of ASD in a task of association between unfamiliar objects and pseudowords. Significant differences were found between ASD (n = 57) and TD (n = 57) Spanish speaking toddlers in the number and time of fixation. The TD children showed more and longer fixations on eyes and mouth while the ASD children attended almost exclusively to objects, making it difficult to integrate lexical and phonological information. Moreover, the TD toddlers looked at the mouth when the pseudoword was produced while the ASD toddlers did not. Gaze fixation on eyes and mouth during word learning recorded by eye-tracking may be used as a biomarker for the early diagnosis of ASD.

The Unique Cost of Human Eye Gaze in Cognitive Control: Being Human-Specific and Body-Related?

This study investigated the eye gaze cost in cognitive control and whether it is human-specific and body-related. In Experiment 1, we explored whether there was a cost of human eye gaze in cognitive control and extended it by focusing on the role of emotion in the cost. Stroop effect was found to be larger in eye-gaze condition than vertical grating condition, and to be comparable across positive, negative, and neutral trials. In Experiment 2, we explored whether the eye gaze cost in cognitive control was limited to human eyes. No larger Stroop effect was found in feline eye-gaze condition, neither the modulating role of emotion. In Experiment 3, we explored whether the mouth could elicit a cost in Stroop effect. Stroop effect was not significantly larger in mouth condition compared to vertical grating condition, nor across positive, negative, and neutral conditions. The results suggest that: (1) There is a robust cost of eye gaze in cognitive control; (2) Such eye-gaze cost was specific to human eyes but not to animal eyes; (3) Only human eyes could have such eye-gaze costs but not human mouth. This study supported the notion that presentation of social cues, such as human eyes, could influence attentional processing, and provided preliminary evidence that the human eye plays an important role in cognitive processing.

Research-Based Intervention (RBI) for Autism Spectrum Disorder: Looking beyond Traditional Models and Outcome Measures for Clinical Trials

The rising prevalence of Autism Spectrum Disorders (ASD) has led to a quickly increasing need for effective interventions. Several criteria and measures have been developed to critically assess these interventions with particular focus on the evaluation of the efficacy. Given the huge diversity of ASD symptoms and the different levels of severity across individuals, identifying a one size fits all intervention approach is challenging, and the question What works and for whom? Remains still unanswered. Why do we seem to be dragging our feet on this fundamental issue? The main aim of this paper is to answer this question through four non-alternative points. First, there are a scarce number of studies with a solid methodology. Secondly, most trials on intervention efficacy for ASD are designed exclusively in terms of behavioral outcomes. Thirdly, there is a reduced use of biologically oriented outcome measures. Fourthly, in most clinical trials, appropriate practices emerging from research evidence are not systematically applied. A strong effort to improve the methodology of clinical trials is mandatory for the future of autism research. The development of a research-based intervention (RBI) perspective aimed at better integrating: (a) evidence-based approaches; (b) more sensitive behavioral outcome measures; and (c) biomarkers, with the aim of increasing a more detailed clustering of phenotypes, may strongly improve our approach to a precision medicine.