Invention:
This technology is a text data set of characteristics of Autism Spectrum Disorder (ASD), consisting of “human-labeled” and “machine-labeled” data. The dataset consists of fragments of clinical text describing ASD typical behaviors in children. This data set will be used by algorithms for a number of applications. One application would be to serve as a training system for clinicians to better diagnose and recognize ASD in young children. Another application, for a more commercial audience, would be to use this algorithm in Alexa-like devices to allow parents to match behaviors of their children to receive a correct or appropriate diagnosis.
Background:
Currently, there are standardized algorithms used for diagnosing Autism Spectrum Disorder (ASD) in children. However, this technology claims to have the largest set of examples in the text data set. The utilization of such a large data set could be imperative to progressing the accuracy of ASD testing and diagnosis in young children. The technology also has clinical and commercial applications, allowing technicians to train and better recognize signs of ASD in children at varying stages and with more specificity, while also allowing parents to monitor their children in their natural environments and readily compare their behaviors to a catalog of examples and clinical text to present to their clinician.
Applications:
Advantages:
- Accessible
- Vast data set
- Easy to integrate
