Autism AI
Smarter Early Autism Screening
The Autism AI Project represents a significant step forward in autism research and diagnosis. Through innovation, collaboration, and rigorous evaluation, we aspire to create a transformative tool that enhances autism detection and supports families worldwide.
.
Overview
Autism Spectrum Disorder (ASD) is a lifelong condition that affects social interaction, communication, and behaviour. In New Zealand, children are typically diagnosed around the age of 6 or 7, even though signs of autism can appear as early as 18 months. Early diagnosis is crucial, as it enables children and their families to access the necessary support and resources to improve their quality of life. However, the current screening and diagnosis process presents significant challenges. Current methods are time-consuming, costly, and often inaccessible, particularly for underserved communities such as Māori and Pasifika. Many families face long waiting times, limited access to specialists, and financial barriers, all of which can delay vital early interventions.
Methodology
Our research aims to address these screening challenges while uncovering and mitigating algorithmic biases that may arise from data imbalances or model limitations. To achieve this, we developed the Ensemble Artificial Intelligence Platform Process (EAIPP), an advanced novel approach that integrates supervised, unsupervised, and semi-supervised AI techniques to enhance diagnostic accuracy and reliability.
Objectives
Development
Creating an advanced AI-powered screening tool to improve early autism detection and diagnosis.
Detection
Enhancing Autism identification by integrating supervised, unsupervised, and semi-supervised AI techniques.
Evaluation
Assessing the accuracy, reliability, and usability of the Autism AI system through trials and real-world applications.
Equity & Accessibility
Ensuring that Autism screening tools are accessible and beneficial to all communities, including underserved populations such as Māori and Pasifika.