A tablet-based application called SenseToKnow has the potential to address disparities in autism spectrum disorder (ASD) diagnoses across different genders, races, and ethnicities, according to a study funded by the National Institutes of Health (NIH). Developed by researchers from Duke University’s Autism Center of Excellence, the SenseToKnow app offers a screening tool for toddlers to detect ASD, possibly replacing parent questionnaires commonly used for diagnosis, which tend to miss cases, particularly in girls and children from diverse backgrounds.
In contrast to traditional questionnaires, the SenseToKnow app engages toddlers in watching short films and records and analyzes their behavioral responses to onscreen stimuli. This includes changes in facial expressions, blink rate, head movement, attention span, and more. In the NIH-backed study, 475 toddlers between 17 months and 3 years old used the SenseToKnow app during routine well-child visits.
Out of the participants, 49 children were eventually diagnosed with autism. The study revealed that the app identified these children with a sensitivity of 87.8% and accurately ruled out those without autism with a specificity of 80.8%.
For participants who screened positive for autism using the app, there was a 40% probability of receiving an official diagnosis, compared to a 15% probability rate for those who screened positive with the standard questionnaire. Combining the results of the app and parent survey increased the accuracy further, resulting in a 63.4% likelihood that a positive screening would lead to an official diagnosis.
The study’s findings were consistent across genders, races, and ethnicities. However, specificity was lower among Black children compared to those of other races, with a specificity rate of 54% versus about 85%.
While further research into the app’s performance across various demographic groups is underway, the initial results hold promise in improving the accuracy of autism screening and reducing disparities in access to diagnosis and intervention. The app is seen as a valuable addition to existing autism screening questionnaires.
The researchers emphasized that the effective use of such screening tools requires training and systematic implementation by healthcare providers. Positive screening results should be followed by appropriate referrals and services to ensure timely access to interventions that can influence long-term outcomes.
Duke University researchers are not alone in developing digital tools for enhancing autism diagnosis rates. Cognoa received FDA clearance for an app that uses machine learning technology and videos of children to aid in diagnosing autism in kids aged 18 months to 6 years. EarliTec Diagnostics also gained FDA approval for its eye-tracking technology, which assists in diagnosing toddlers aged 16 months to two and a half years. These advancements in digital tools aim to improve the early identification and management of autism spectrum disorder.