Saturday 08 March 2025
Researchers have developed a new tool that can accurately assess motor imitation skills in children with autism spectrum disorder (ASD). This skill is crucial for social learning and development, but is often impaired in individuals with ASD.
The traditional method of assessing motor imitation involves observing a child’s ability to mimic actions performed by an adult. However, this approach has several limitations, including the need for extensive human training and labor-intensive data normalization processes.
To address these challenges, scientists have created a computerized assessment tool called CAMI-2DNet. This innovative system uses deep learning algorithms to analyze video recordings of children performing simple motor tasks, such as clapping or waving. By analyzing these actions, the system can accurately identify which body parts are being used and how they are being moved.
The CAMI-2DNet system consists of three main components: a motion encoder, a body encoder, and a view encoder. The motion encoder extracts features from the video footage that capture the movement patterns of the child’s limbs, while the body encoder identifies the specific joints and segments involved in the action. Meanwhile, the view encoder disentangles the viewpoint-related features to ensure that the system is not biased by camera angles or lighting conditions.
The system then uses these extracted features to calculate a motor imitation score, which reflects the similarity between the child’s actions and those of an adult performing the same task. This score can be used to identify children with ASD who exhibit impaired motor imitation skills, as well as to monitor progress over time in response to interventions or therapies.
The researchers tested CAMI-2DNet on a dataset of 46 children with ASD, comparing its performance to other existing assessment tools. The results showed that CAMI-2DNet was able to accurately identify children with ASD and distinguish them from neurotypical children, outperforming traditional methods in terms of diagnostic accuracy.
One of the key advantages of CAMI-2DNet is its ability to provide a detailed breakdown of which body parts are contributing to the child’s motor imitation difficulties. This information can be used by clinicians to develop targeted interventions that address specific areas of impairment, potentially leading to more effective treatment outcomes.
The development of CAMI-2DNet represents an important step forward in the assessment and diagnosis of ASD, offering a more efficient and accurate tool for clinicians working with children on the autism spectrum.
Cite this article: “Accurate Motor Imitation Assessment Tool Developed for Children with Autism Spectrum Disorder”, The Science Archive, 2025.
Autism Spectrum Disorder, Motor Imitation Skills, Children, Assessment Tool, Computerized System, Deep Learning Algorithms, Video Recordings, Body Parts, Diagnostic Accuracy, Targeted Interventions
