EarlySee: Autism Screening in Infancy and Toddlerhood using Smartphones

Project Description:

Background: Autism is a neurodevelopmental disorder, and 1.5% of children (one out 68 children) in the United States are diagnosed with autism yearly. The current screening for autism highly depends on the subjective judgment of parents and primary medical doctors. Specifically, screening for autism relies solely on observations conducted during a routine well visit, and the so-called "wait and watching" screening convention often results in a 9-24 month delay in pursuing an autism diagnostic assessment. It is an urgent demand to eliminate the current approach of “watch and wait” and move toward a rapid, precise, and objective screening method for autism. Goal: this project is to develop such a cost-effective, widely-accessible, and quantitative tool, i.e., a smartphone app called EarlySee, through multi-dimensional analysis (e.g., face expression, gaze, etc), which enables (a) a time-efficient screening (i.e., a 40-second process), (b) shortening the amount of time necessary to receive a diagnosis, (c) providing a more natural screening environment for children and (d) beginning early intervention services of autism. This transformative screening solution of autism is expected to promote timely diagnosis and intervention for autism, which enables that children with autism can gain significantly better progress and eventually become indistinguishable from same-aged peers with typical development. EarlySee can provide the risk estimation of Autism towards children as young as 12 months old.

EarlySee User Privacy Policy: detailed link

Disclaimer: The EarlySee Project was launched since March 2014. The team didn't receive any grant support, and the project is supported by a group of passionate researchers who wants to promote the childhood mental health.

Updates & News:

  • 2018/04 EarlySee is reported by multiple high-impact international media, including CNN, Fox8 and ACM's News.
  • 2018/04 [The new version is updated on April 4th] The android app of EarlySee is ready to use on April 2nd, World Autism Awareness Day. See the news [link]. If you want to screen your child, please find the EarlySee app in Google Play Store. We will launch the iOS version soon.
  • 2018/02 The android app of EarlySee will be launched around early April, 2018.
  • 2017/04 The EarlySee project is selected as one of the finalists (8 out of 130+) in Global Vodafone Wireless Innovation Competition. [link]
  • 2016/11 The EarlySee technology is in the news. [link]


    Wenyao Xu (PI) - contact: wenyaoxu@buffalo.edu

    Chen Song (Ph.D. student)

    Matthew Stafford (UG student)

    Kun Woo Cho (UG student)

    Tri Vu (UG student)


    Dr. Michelle Hartley-McAndrew, Women and Children Hospital of Buffalo, USA (2014 - 2017)

    Dr. Kathy Doody, Buffalo State University, USA (2014 - 2017)

    Dr. Feng Lin, University of Colorado, Denver (2015 - present)

    Sennotech Group (创感科技), China (link) (2017 - present)

    Related Publications:

  • [3] Tri Vu, Hoan Duc Tran, Kun Woo Cho, Chen Song, Feng Lin, Michelle Hartley-McAndrew, Kathy Doody, Chang Wen Chen, Wenyao Xu, "Efficient and Effective Visual Stimuli Design for Quantitative Autism Screening: An Exploratory Study", IEEE International Conference on Biomedical and Health Informatics (BHI'17), Orlando, Florida, February 2017
  • [2] Chen Song, Aosen Wang, Kathy Ralabate Doody, Michelle Hartley-McAndrew, Jana Mertz, Feng Lin, Wenyao Xu, "A Comparative Study of Social Scene Parsing Strategies between Children with and without Autism Spectrum Disorder", North American Journal of Medical Sciences (NAJMS), Volume 9, Number 3, July 2016, Pages 96 - 103
  • [1] Kun Woo Cho, Feng Lin, Chen Song, Xiaowei Xu, Michelle Hartley-McAndrew, Kathy Doody, Wenyao Xu, "Gaze-Wasserstein: A Quantitative Screening Approach to Autism Spectrum Disorder", IEEE Annual Wireless Health Conference (WH'16), Bethesda, MD, October 2016