June, 2018

Select a different month in the archive

Face-recording App May Screen for Autism

By Chelsea Toledo, M.A. on June 25, 2018


Background: Recent years have seen a huge surge in the number of apps created for Autism Spectrum Disorder (ASD), with the current number in the Apple app store topping 1,000. Many of these apps are aimed at instilling academic and functional skills in individuals with ASD. Others help users find support and services or provide news and raise awareness about ASD.


What’s New: A new study explored the use of an app for screening children with ASD. The researchers developed an iPhone app to detect differences in emotions and attentiveness – two areas affected in ASD. The app recorded the reactions of children as they watched movies on an iPhone, and then coded their behaviors. In all, 1,756 families with children between the ages of 1 and 6 years old participated over the course of a year, resulting in 4,441 videos of children in their natural settings. In addition, 5,618 demographic surveys were received from caregivers.


The researchers found:

  • 32 percent of the participants were high risk for ASD, as defined by either having a caregiver report an ASD diagnosis having a high score on the M-CHAT ASD screening.
  • Video clips - which included bubbles, bunnies, mirrors, and toys and songs - were each watched by more than half of the participants.
  • Girls who were high risk for ASD demonstrated significantly less attention to the videos on bubbles, bunnies, and mirrors than girls who were low risk for ASD. This association was not the same for boys.
  • Children with high M-CHAT scores were less likely to have positive emotional reactions to the bubbles movie than their peers with lower scores.


Why it’s important: This study suggests that app-based method could one day be leveraged to help screen for ASD in children. Because the screening would take place in children’s home environments, the results could be more meaningful than in-office interactions.


Image Credit: Kaethe Richter Henning, University of New Mexico, downloaded from the NIH Image Gallery under a Creative Commons license

Help me understand :

Source(s) :

Reactions to Robot Could Reveal Internal Patterns in Autism

By Chelsea Toledo, M.A. on June 7, 2018


Background: Biologically speaking, a genotype is an organism’s underlying genetic identity, made up of inherited factors. A phenotype refers to observable characteristics such as social behavior. An intermediate term endophenotype is used to define stable, measurable traits, observed in multiple individuals, with a clear link to underlying genetic makeup.


What’s New: A new study assessed whether artificial intelligence could help reveal endophenotypes of individuals with Autism Spectrum Disorder (ASD).  The researchers leveraged a female android robot, known as ACTROID-F, to interact with a total of 46 individuals representing four groups – adolescents (ages 13-17) with ASD, typically developing adolescents, young children (ages 1-5) with ASD, and typically developing young children. The robot was programed to smile, nod, move her eyebrows, and demonstrate other evidence of humaneness while interacting with the participants as they completed various tasks. The researchers administered a questionnaire to the adolescent participants and to the parents of the young children, measuring the degree to which the participants felt that the android displayed humaneness (i.e., whether their impression of her was natural, positive, competent, kind, conscious, human-like, responsible, nice, friendly, intelligent, and biological). They found that scores varied significantly between the participants with ASD and their typically developing peers.


Why it’s important: This study suggests that artificial intelligence could reveal an endophenotype for people with ASD. Further research could provide more information about the underlying genetic causes for these patterned differences.

Help me understand :

Source(s) :