Using R Shiny to Diagnose Autism Spectrum Disorder

Free State Analytics built an application to assist behaviorial analysts who test children for autism.

Unfortunately, most methods for measuring language variability are difficult to calculate. Data visualizations, statistical tests, and sorted tables help behavioral analysts see the big picture, but are time consuming to produce.

A child with autism will typically show an imbalance in language development. Two psychologists specializing in autism spectrum disorder, Dr. Lee Mason and Dr. Alonzo Andrews, developed the Vox Analysis framework, which uses play therapy to collect data and measure the child’s language development.

Dr. Mason and Dr. Andrews partnered with Free State Analytics to build a Shiny application that would run these calculations in a minute.

Why Free State Analytics?

Dr. Mason and Dr. Andrews needed an app that would allow for:

  1. User authentication
  2. Dynamic input generation
  3. Downloading and re-uploading evaluation data
  4. Rules to ensure only valid entries
  5. Statistical functions and data visualizations

“Off-the-shelf” tools can make it difficult to deliver requirements as advanced as these. Free State Analytics prefers to use code-based solutions such a R Shiny (and others) to ensure the precision that’s required when it comes to project delivery.

Custom Shiny Input Development

One of the most innovative parts of this project was the development of a matrix-like input for Shiny. The user defines the number of stimuli from a play therapy session, labels them, and then inputs a 1 or 0 based on the child’s response in that particular language test. This involved building a series of custom functions and tweaks to formatting to match the shiny.semantic package’s look and feel.

Need Your Own Custom Application

As we like to say, we want the hard projects.

Have your own project? Contact us today.