A team of researchers has recently developed a new diagnostic tool for autism that analyzes how the patient observes faces. Current methods of diagnosing autism include surveys and evaluations from psychologists, but with this new technology, physicians may be able to identify the condition through one’s visual gaze. This study was led by scientists from the University of Waterloo and was recently published recently in the journal Computers in Biology and Medicine.
Those with autism display impaired social skills that lead to abnormalities in eye contact. These Waterloo researchers hypothesized that by presenting such patients with a facial image and recording their visual gaze, they may be able to detect the condition through these abnormalities.
Background of the Eye Movement Study
The diagnostic tool was developed through evaluation of 17 children with autism and 23 without. The average age of each group was 5.5 and 4.8 years, respectively. To gather data regarding each participant’s facial analysis, they were shown 44 images of faces on a screen. An infrared eye-tracking system was incorporated into this process to identify and interpret the locations where each child directed their attention using emission and reflection of wave from the eye.
Each image was divided into specific areas of interest (AOIs) where the children directed their focus, including the right and left eyes, areas beneath them, mouth, nose, and other screen regions. The researchers aimed to find both how the participants moved their eyes to scan faces, or their saccades, and how much time they spent focused on each AOI.
These factors were identified using four concepts from network analysis, the first being the quantity of AOIs the child shifts focus to from one specific AOI. The second concept accounted for how frequently an AOI is involved during saccades between two other AOIs. The third and fourth concepts focused on the speed of saccades and the importance of an AOI based on the transitions between it and other significant points.
The researchers found that those with autism spend significantly more time looking at the mouth and less at the eyes compared to those with typical development. They also noted that those without autism displayed faster saccades than the children with the condition, and that autism gaze patterns were best characterized through betweenness centrality.
Implications of this Tool for Autism Diagnosis
The team feels that their technique makes diagnosing autism much less stressful for young patients and that it will help prevent false diagnoses if used alongside existing methods.
“Many people are suffering from autism, and we need early diagnosis especially in children,” said Mehrshad Sadria, a master’s student in the University of Waterloo’s Department of Applied Mathematics. “The current approaches to determining if someone has autism are not really child-friendly. Our method allows for the diagnosis to be made more easily and with less possibility of mistakes.”
Sadria went on to add that their technology can be used in diagnosing all patients with autism, but that the team sees it being best fit for use in children.
“It is much easier for children to just look at something, like the animated face of a dog, than to fill out a questionnaire or be evaluated by a psychologist,” added Anita Layton, who supervises Sadria and teaches Applied Mathematics, Pharmacy and Biology at Waterloo. “Also, the challenge many psychologists face is that sometimes behaviors deteriorate over time, so the child might not display signs of autism, but then a few years later, something starts showing up. Our technique is not just about behavior or whether a child is focusing on the mouth or eyes. It’s about how a child looks at everything.”
University of Waterloo researchers discover a new way of detecting autism in children — by studying how they scan pictures of faces. https://t.co/D0QAwTagUu!
— Elizabeth Payne (@egpayne) July 9, 2019