The ocular fundus is the interior surface of the eye, and an ophthalmologist can learn a lot about a patient’s health by examining it. However, there’s a problem: an ocular fundus camera can’t capture a useful image unless the eye is brightly lit, but this makes the pupil constrict, obstructing the camera’s view. Ophthalmologists use pupil-dilating eye drops to block the eye’s response to light, but these are uncomfortable and can cause blurred vision for several hours. Now, researchers at the University of Illinois at Chicago have built a Raspberry Pi-based fundus camera that can take photos of the retina without the need for eye drops.
Dr Bailey Shen and co-author Dr Shizuo Mukai made their camera with a Raspberry Pi 2 and a Pi NoIR Camera Module, a version of the Camera Module that does not have an infrared filter. They used a small LCD touchscreen display and a lithium battery, holding the ensemble together with tape and rubber bands. They also connected a button and a dual infrared/white light LED to the Raspberry Pi’s GPIO pins.
When the Raspberry Pi boots, a Python script turns on the infrared illumination from the LED and displays the camera view on the screen. The iris does not respond to infrared light, so in a darkened room the operator is able to position the camera and a separate condensing lens to produce a clear image of the patient’s fundus. When they’re satisfied with the image, the operator presses the button. This turns off the infrared light, produces a flash of white light, and captures a colour image of the fundus before the iris can respond and constrict the pupil.
This isn’t the first ocular fundus camera to use infrared/white light to focus and obtain images without eye drops, but it is less bulky and distinctly cheaper than existing equipment, which can cost thousands of dollars. The total cost of all the parts is $185, and all but one are available as off-the-shelf components. The exception is the dual infrared/white light LED, a prototype which the researchers explain is a critical part of the equipment. Using an infrared LED and a white LED side by side doesn’t yield consistent results. We’d be glad to see the LED available on the market, both to support this particular application, and because we bet there are plenty of other builds that could use one!
Read more in Science Daily, in an article exploring the background to the project. The article is based on the researchers’ recent paper, presenting the Raspberry Pi ocular fundus camera in the Journal of Ophthalmology. The journal is an open access publication, so if you think this build is as interesting as I do, I encourage you to read the researchers’ presentation of their work, its possibilities and its limitations. They also provide step-by-step instructions and a parts list to help others to replicate and build on their work.
It’s beyond brilliant to see researchers and engineers using the Raspberry Pi to make medical and scientific work cheaper and more accessible. Please tell us about your favourite applications, or the applications you’d develop in your fantasy lab or clinic, in the comments.