Hardware components | ||||||
 |
| × | 1 | |||
| × | 1 | ||||
The goal was create a simple proof-of-concept for detecting diabetic retinopathy in the field with inexpensive hardware.
The project leveraged Jetson Nano with Jetpack, a USB camera and existing public domain retinal images to train simple image recognition model in a Jupyter notebook.
The model was trained with 8 images for each of the five categories of the disease, and the results were very encouraging.
The simple setup used a laptop to display the images and captured them using a webcam. The next step for me is to create an eyewear device that is capable of capturing the retinal images. There is some past research work from leading universities that has explored the topic.
Overall setup
The laptop was used to display the retinal images. Those were captured with the webcam (mounted on the box that the Nano came in!). The images were captured using the Jupyter notebook, and used the train the model. Then the trained model was used to evaluate additional images from the laptop. Eventually the laptop/webcam will replaced by a small wearable eyepatch that will capture the retinal image.
Comments