TRACERY ARTIFICIAL INTELLIGENCE
Incorporates Dedicated Advanced Computing Algorithms Including, Machine Learning & Artificial Intelligence for Multi-Modal , Cross-Modal Image Analysis
Tracery's image analysis solution includes;
pre-processing modules for image export, quality assessment, de-identification, and cross-modal registration of images from multiple devices
Tracery’s algorithms were trained by domain experts, including retina specialists with a specific interest in AMD.
​
Cross-sectional and longitudinal data were obtained in two REB approved natural history studies.
Human graders used Tracery's active interface to train ML/AI algorithms for en face and cross-sectional image analysis.
Region of interest analysis was performed to highlight risk factors for disease progression.
Our algorithms benefit from data augmentation
Tracery has developed dedicated neural networks to automatically align and analyze de-identified patient images, and deploys multiple autoencoders & deep voting to increase the accuracy of image segmentation output accuracy