This tool automatically classifies retinal images into three categories: Normal, ASD, and ADHD. You can also view a heatmap visualization to see which areas of the image contributed to the classification.
Steps to Use
Browse through the available sample images and select one by clicking on it.
Click the Predict button to send the image to the model.
Wait for the model to process the image and display the results.
Understanding the Output
The tool provides:
Predicted Class: The model screening prediction.
Confidence Score: The model confidence in that prediction.
GradCAM Overlay: A heatmap that highlights which regions of the image contributed to the prediction.
Confidence Distribution: Distribution of previous model confidence outputs for that condition, which can be used to contextualize the model's confidence and infer reliability of results.
Notes
This tool is for research purposes and should not be used for medical diagnosis.
Part of "In Vitro and In Silico Retinal Models of Neurodevelopmental Disorders."