Ear AI Demo Guide

1. Go to https://demo.belle.ai and select the AI module.

2. Upload a corresponding image. You will have the option to add multiple images (look for the plus sign that will appear after the first image is uploaded).

3. The AI will then analyze the image, outlining the image’s key features including: the eardrum, wax, and other abnormalities in the external auditory canal (containing foreign bodies, granulomas, and redness); the middle ear (including fluid, air bubbles, and swelling); and the tympanic membrane (covering perforations, blood, and scarring). These features may differ between images (ei. Redness or Fluid may be outlined if present).

• The Irrelevant Score indicates the image’s degree of relevance to the disease the model has been trained on (ie. ensures an image of the ear has been uploaded).
• The Image Quality Score evaluates image blur and sharpness. The larger the score, the better (0.0 - 1.0).
• Main Suggestions indicates the AI’s predicted probability of particular conditions. The associated Belle Index Matching (BIM) scores indicate how much a given image compares to other images of disease classes or subclasses. The higher the BIM score, the stronger the correlation between the given image and the database images for a particular condition.
• The AI will tell if there is a possible infection or not. We have a list of infection diseases. If one of them appears in the list (BIM score higher than 0.2), the AI will return infection positive.
• You may add notes or comments under “What do you think about the result?”. You may check the box of what condition you believe is present. Your input will not change any scoring.

4. You may revisit previous analyses via the History tab (located on the sidebar) which automatically saves previous results (yet comments on such analyses must be saved manually by clicking the save button).

Questions or Feedback?