A recent study using functional near-infrared spectroscopy compared neural responses of preschool children with and without autism to videos presenting a human and a humanoid robot. Neural activity of children with autism differed in the situation when they were interacting with a video containing a humanoid robot compared to interacting with a video containing a human being. Neural activity of children without autism, in contrast, was similar in these two situations.
The study was published in the International Journal of Psychophysiology.
Autism or autism spectrum disorder (ASD) is one of the most common disability in children. In the United States, around one in every 50 school-aged children is diagnosed with the autism spectrum disorder. Children with autism have problems in communicating with others and with social interaction. Compared to other children, they have difficulty sharing feelings and expressing interests, identifying intentions of others, and providing appropriate responses.
In recent decades, scientists started using neuroimaging techniques to investigate brain mechanisms underlying the atypical responses in social interactions of children with ASD. Quite a few of these studies showed that the brains of children with autism show different activity when interacting with other people compared to their peers without autism.
Studies have shown that children with autism show great interest in humanoid robots. One study revealed that these children show “concerted attention to toys or objects that they like, but have difficulty sharing attention or interests with other people.” Also, when choosing between a toy truck and a humanoid robot, autistic children are tend to show greater interest in the humanoid robot.
Study author Sheumeng Hou of the Harbin Institute of Technology and his colleagues wanted to explore whether the brain activation patterns of young autistic children differed when they interact with a humanoid robot compared to situations when they interact with a human. They also wanted to know whether these activation patterns are specific for children with autism or are found in children not diagnosed with autism as well.
“We sought a neurobiological basis for the atypical responses of young autistic children when interacting with a robot, thereby providing support for using robots as effective methods in clinical settings,” the researchers explained.
The study included 45 children diagnosed with autism spectrum disorder (41 males) and 53 children without autism (36 males). They were between 4 and 6 years of age. Children participated in the experiment along with their parents.
In the experiment, a child was seated in front of a desk upon which a screen was placed. The parent set behind the child and the experimenter was in the same room. A cap for taking functional near-infrared spectroscopy images was placed on the child’s head.
For this, researchers used a continuous-wave NIRSport system with a sampling rate of 3.47 Hz measuring at two wavelengths (760 nm and 850 nm). This system measures brain function by monitoring changes in relative concentrations of oxygenated and deoxygenated hemoglobin. In this study, it was placed to record left and right dorsolateral prefrontal cortex regions of the brain.
Children were shown a series of 12 video clips based on daily communication scenarios in which the character in the clip would pose a question or say something and then make a pause (during which the child watching the clip could respond). The children were instructed to react to the videos “in the way they like.”
Four of the videos presented a human talking to the child (human condition), 4 contained a humanoid robot (robot condition) and 4 contained squares shown instead of a speaker (square condition). Videos were shown in randomized order to each child and they were in the Mandarin Chinese language.
Results showed that, in children with autism, neural activity in the dorsolateral prefrontal cortex region of the brain was lower when they viewed clips with a robot than when they were shown clips with a human. Their neural activity in this part of the brain when viewing clips with a robot was also lower than neural activity of children without autism when they were viewing the same type of clips.
Additionally, children with autism who showed higher neural activity when viewing clips with robots tended to show lower neural activity when viewing clips with humans and vice versa. In contrast to this, children without autism who showed higher neural activity when watching videos with a human also showed higher neural activity when watching videos with a robot.
“While neurotypical children showed comparable neural activity to humanoid robots and human beings, the children with ASD showed significantly different neural activity under those two conditions,” the researchers wrote. “Children with ASD may need more selective attention resources for human interaction than for robot interaction. It is also much more difficult for children with ASD to neglect the attraction of robots.”
The study provides a valuable contribution to knowledge about the neural mechanisms of autism. However, it should be noted that the study used video clips and reactions to real social interactions might not be the same. Additionally, it focused solely on neural activation and did not record how engaged the children were with the video clips.
The study, “Young children with autism show atypical prefrontal cortical responses to humanoid robots: An fNIRS study”, was authored by Shumeng Hou, Ning Liu, Jun Zou, Xuejiao Yin, Xinyue Liu, Shi Zhang, Jiesheng Chen, Zhen Wei.
