Rather than interviewing subjects or administering questionnaires, Bappaditya Mandal and his colleagues used an AI program to study the body movements of autistic children in YouTube videos. However, the children and their parents did not opt in to having their home videos used for scientific research.
In the summer of 2019, a behavioral study on children with autism was published by a team at England’s Keele University. As a highlight of their study, Bappaditya Mandal and his colleagues trained an artificial intelligence to study the body movements of children with autism in online YouTube videos. They then used it to classify their behaviors as either typical or atypical. As reported in The Atlantic, “The researchers’ goal…is to use computers to more quickly evaluate edge cases that might normally require lab equipment or invasive tactile sensors.”
The AI’s algorithms track the appearance of tremors or seizures in children with epilepsy, as epilepsy and autism often go hand in hand. This video analysis is designed to help families and scientists establish understanding on when behaviors appear, what triggers them and which parts of the body are most affected. These are “all the things the doctors need to know in order to do a good diagnosis,” Mandal explained.
The team sought out YouTube videos, uploaded sometimes by parents or autism advocacy groups, to compile a database and repurpose the clips as valuable analytical data. Mandal notes that although children’s faces appear in the database, the software doesn’t scan their faces or identify them; it just uses machine learning to read their body language. Regardless, there was no consent given by the children or their parents.
This is all part of the growing field of study called digital phenotyping, where social media and smartphone data are used to study someone’s health status and dynamics. And in today’s digital age, social media offers as much potential data as it does potential privacy risks.
But this is a complicated trade-off; just because it’s public information, doesn’t make it ethical to utilize. “Digital phenotyping only deepens this complexity, creating new risks and opportunities to consider as we try to measure the costs and benefits of our ever more digital lives,” reports The Atlantic.
Call to Action: To learn more about the risks and benefits of digital phenotyping, sign up for Trialsite News’ newletter. We work to cover all possible information that can both inform our readers and help them in their personal healthcare decision-making. And if you’d like to contact the researchers involved, we offer concierge services that can help you get in touch.