Biomedical · Diagnostic Imaging & Analysis
Retinal Fundus Image Analysis for Diabetic Retinopathy
Build a imaging rig (smartphone + lens) to capture retinal-style images, then run an AI classifier. Combines real optical hardware with computer vision. (Use model eyes/public images ethically.) NOTE: Educational project only. NOT a medical device and not for clinical or diagnostic use.
Major
Biomedical
Focus area
Diagnostic Imaging & Analysis
Total hours
40-70 hours
What you'll need
- 20D condensing lens for fundoscopy
- smartphone clip-on macro lens kit
- LED ring light for photography
- PLA 3D printer filament 1.75mm for mounts
- Raspberry Pi Camera Module 3 12MP
As an Amazon Associate, I earn from qualifying purchases.
Steps
- 3D print a mount that holds the lens, light ring, and camera in line.
- Capture images of a model eye and practice getting clear, focused shots.
- Download a public retinal image dataset and train or fine-tune an image classifier.
- Test the classifier on held-out images and measure accuracy.
- Run your own captured images through the classifier and discuss the limits.
- Document the rig, model, and results for your portfolio.
What to photograph for your portfolio
- Imaging rig build
- captured sample images
- AI detection results
- accuracy metrics
Resume bullet starters
Copy one, then swap in your own numbers.
Engineered a retinal fundus image analysis for diabetic retinopathy using Biomedical imaging, optics, deep learning
Applied computer vision and hardware integration to implement, test, and validate the analysis end-to-end
Quantified performance with [insert your result, such as accuracy, error reduction, response time, or load capacity] after iterative tuning