UCLA researchers use deep learning to turn a smartphone into a lab-quality microscope.
Got a science lab in your pocket? Probably not, but it may soon be a reality—and one that could benefit low-income areas that lack medical resources. Researchers at the UCLA Samueli School of Engineering used deep learning, a branch of artificial intelligence (AI), to develop a technique that can recognize and enhance microscopic details in photos taken by smartphones and turn them into quality lab-grade images down to a scale of approximately one millionth of a meter.
“Using deep learning, we set out to bridge the gap in image quality between inexpensive mobile phone-based microscopes and gold-standard bench-top microscopes that use high-end lenses,” said Aydogan Ozcan, chancellor’s professor of electrical and computer engineering and bioengineering. “We believe that our approach is broadly applicable to other low-cost microscopy systems that use, for example, inexpensive lenses or cameras, and could facilitate the replacement of high-end bench-top microscopes with cost-effective, mobile alternatives.”
Smartphone attachments that can capture microscope-scale images have been around for some time. Unfortunately, the cameras aren’t made for high-resolution microscopic images. Still, an attachment was a good starting place for the UCLA team, which developed one that could be 3D printed and placed over a smartphone lens to increase the resolution and visibility of minute details in images. While the technique worked, it still couldn’t compete with high-end lab equipment.
The researchers then turned to AI to make up the difference. They took images of lung tissue samples, blood and Pap smears with a standard laboratory-grade microscope, and then with a smartphone that had the 3D-printed microscope attachment. The pairs of corresponding images were fed into a computer system with a UCLA-developed deep-learning-based computer code to “learn” how to rapidly enhance the lower resolution images. The team’s technique distinguishes and enhances microscopic details in an image taken by a smartphone, resulting in improved image color and resolution.
This development could have far-reaching implications. Since it uses attachments that can be 3D printed for less than a $100 each, resource-poor areas could gain access to high-quality medical diagnostics at a fraction of the cost.
Interested in more ways to use a smartphone as a tool? Check out Gadget Converts Smart Phones into Laser Rangefinders and Android App Transforms Hand-Drawn Sketches into 3D Building Models.
