Sign In

Communications of the ACM

ACM TechNews

UCLA Engineers Use Deep Learning to Reconstruct Holograms and Improve Optical Microscopy


The technique uses deep learning to produce high-resolution pictures from lower-resolution microscopic images.

University of California, Los Angeles researchers have developed two new applications for machine learning.

Credit: Ozcan Research Group/UCLA

Researchers at the University of California, Los Angeles (UCLA) have developed two new uses for machine learning--one for reconstructing a hologram to form a microscopic image of an object, and another for improving optical microscopy.

The researchers say the new holographic imaging techniques produce better images than conventional methods using multiple holograms, and deployment is easier because it requires fewer measurements and performs computations faster.

In the first study, the researchers produced holograms of Pap smears, which were analyzed by a neural network so it could learn to extract and separate the features of the true image of the object from undesired light interference and from other physical byproducts of the image reconstruction process.

"This is an exciting achievement since traditional physics-based hologram reconstruction methods have been replaced by a deep learning-based computational approach," says UCLA's Yair Rivenson.

In the second study, the researchers used the same deep-learning framework to improve the resolution and quality of optical microscopic images.

From UCLA Newsroom
View Full Article

 

Abstracts Copyright © 2017 Information Inc., Bethesda, Maryland, USA


 

No entries found

Sign In for Full Access
» Forgot Password? » Create an ACM Web Account