Advances in neuroimaging, genomic, motion tracking, eye-tracking and many other technology-based data collection methods have led to a torrent of high dimensional datasets, which commonly have a small number of samples because of the intrinsic high cost of data collection involving human participants. High dimensional data with a small number of samples is of critical importance for identifying biomarkers and conducting feasibility and pilot work, however it can lead to biased machine learning (ML) performance estimates. Our review of studies which have applied ML to predict autistic from non-autistic individuals showed that small sample size is associated with higher reported classification accuracy. Thus, we have investigated whether this bias could be caused by the use of validation methods which do not sufficiently control overfitting. Our simulations show that K-fold Cross-Validation (CV) produces strongly biased performance estimates with small sample sizes, and the bias is still evident with sample size of 1000. Nested CV and train/test split approaches produce robust and unbiased performance estimates regardless of sample size. We also show that feature selection if performed on pooled training and testing data is contributing to bias considerably more than parameter tuning. In addition, the contribution to bias by data dimensionality, hyper-parameter space and number of CV folds was explored, and validation methods were compared with discriminable data. The results suggest how to design robust testing methodologies when working with small datasets and how to interpret the results of other studies based on what validation method was used.

The current study investigated whether the amount of autistic traits shown by an individual is associated with viewing behaviour during a face-to-face interaction. The eye movements of 36 neurotypical university students were recorded using a mobile eye-tracking device. High amounts of autistic traits were neither associated with reduced looking to the social partner overall, nor with reduced looking to the face. However, individuals who were high in autistic traits exhibited reduced visual exploration during the face-to-face interaction overall, as demonstrated by shorter and less frequent saccades. Visual exploration was not related to social anxiety. This study suggests that there are systematic individual differences in visual exploration during social interactions and these are related to amount of autistic traits.

Autism is a developmental condition currently identified by experts using observation, interview, and questionnaire techniques and primarily assessing social and communication deficits. Motor function and movement imitation are also altered in autism and can be measured more objectively. in this study, motion and eye tracking data from a movement imitation task were combined with supervised machine learning methods to classify 22 autistic and 22 non-autistic adults. The focus was on a reliable machine learning application. We have used nested validation to develop models and further tested the models with an independent data sample. feature selection was aimed at selection stability to assure result interpretability. Our models predicted diagnosis with 73% accuracy from kinematic features, 70% accuracy from eye movement features and 78% accuracy from combined features. We further explored features which were most important for predictions to better understand movement imitation differences in autism. Consistent with the behavioural results, most discriminative features were from the experimental condition in which non-autistic individuals tended to successfully imitate unusual movement kinematics while autistic individuals tended to fail. Machine learning results show promise that future work could aid in the diagnosis process by providing quantitative tests to supplement current qualitative ones. Autism is a group of complex developmental conditions characterised by deficits in social skills, verbal and non-verbal communication, and restrictive, repetitive behaviours. However, precise expression of symptoms can vary considerably and there are no universal biomarkers. It is one of the most prevalent developmental disorders affecting approximately 1% of the population, resulting in ~700,000 individuals living with autism in the UK 1. Currently, its diagnosis relies on clinical experts using observation, interview, and questionnaire techniques, which depend on interpretative coding. The diagnostic process is complex, long and expensive, and the average waiting time between recognising initial concerns and actual clinical diagnosis is more than 3 years in the UK 2. Thus, valuable time is lost, because early identification and intervention are associated with better outcomes 3. Although, the majority of autistic individuals receive diagnosis in childhood, many remained undiagnosed until adulthood or not at all. The diagnostic process for the adult population is complicated, as current diagnostic instruments have only been validated for use with children 4. With adults, clinicians rarely rely on standardised diagnostic methods making diagnosis less accurate, more subjective and lengthier 5. Thus researching relevant diagnostic criteria for the adult population is critical and listed as one of top ten priorities by the UK's leading autism research charity Autistica 4. In addition to social and communication deficits, current diagnostic criteria recognize repetitive behaviours and movements as core symptoms of autism 6....

This study investigated whether reduced visual attention to an observed action might account for altered imitation in autistic adults. A total of 22 autistic and 22 non-autistic adults observed and then imitated videos of a hand producing sequences of movements that differed in vertical elevation while their hand and eye movements were recorded. Participants first performed a block of imitation trials with general instructions to imitate the action. They then performed a second block with explicit instructions to attend closely to the characteristics of the movement. Imitation was quantified according to how much participants modulated their movement between the different heights of the observed movements. In the general instruction condition, the autistic group modulated their movements significantly less compared to the non-autistic group. However, following instructions to attend to the movement, the autistic group showed equivalent imitation modulation to the non-autistic group. Eye movement recording showed that the autistic group spent significantly less time looking at the hand movement for both instruction conditions. These findings show that visual attention contributes to altered voluntary imitation in autistic individuals and have implications for therapies involving imitation as well as for autistic people’s ability to understand the actions of others.

Autism is a developmental condition primarily identified by social and communication deficits. However, over 70% of autistic individuals also show motor function deficits, which are evident even when simple stereotyped movements are performed. In this study, we have asked 24 autistic and 22 non-autistic adults to perform pointing movements between two markers 30 cm apart as quickly and as accurately as they can for 10 seconds. Motion tracking was employed to collect data and calculate kinematic features of the movement and aiming accuracy. At the group level, the results showed that autistic individuals performed pointing movements slower but more accurately compared to non-autistic individuals. At the individual level, we have used Machine Learning methods to predict autism diagnosis. Nested result Cross-Validation was used, which in contrast to commonly used K-fold Cross-Validation avoids pooling training and testing data and provides robust performance estimates. Our developed models achieved a statistically significant classification accuracy of 71% and showed that even a simple and short motor task enables discrimination between autistic and non-autistic individuals.