ESTEEM Center for Equity in Science, Technology, Engineering, English & Math



Duke’s Autism Screening app: SenseToKnow

How Duke’s autism screening app – SenseToKnow – works
“iPad 2 w/ Smart Cover” by leondel is licensed under CC BY-NC-ND 2.0


According to CDC, one in 36 children are diagnosed with autism. Out of this, many are undiagnosed. This makes it harder for the parents and child to cope with it. Late diagnosis is also a factor of this. However, Duke University and the National Institute of Health (NIH), created an app that is benefitting the field of medicine and technology. This app is known as SenseToKnow.

How does it work?

SenseToKnow is an autism screening app. The app has parents make their child watch a 10-15 minute video while a camera records the child’s behavior, such as their gaze, facial expressions and focus. The test can determine whether the kid has a high chance of becoming autistic. A good screening test will be able to accurately recognize who and who isn’t autistic. These two aspects are known as sensitivity and specificity.

In tests with low sensitivity and high specificity, the majority of positive results will be pretty accurate, but there will be a lot of false negatives. On the other hand, in tests with a high sensitivity but bad sensitivity, then there will be a lot of false positives, and there will not be many false negatives. According to National Geographic, SenseToKnow had a sensitivity rating of 87.8% and a specificity rating of 80.8%.

We need more research

Before using SenseToKnow in primary care, it needs more study. This includes confirming the accuracy of the app in different groups of kids. Geraldine Dawson, the principle investigator in this study, and her collaborators want to test SenseToKnow in a more diverse set of kids because while the accuracy of the app was great, the results weren’t even with all of the groups of patients.


Related Stories:,other%20behaviors%20related%20to%20ASD.

Take Action:



More to Discover