Deep learning improves medical imaging

Deep learning improves medical imaging

Technology News |
By Rich Pell

The researchers used deep learning to both reconstruct a hologram to form a microscopic image of an object, and to improve optical microscopy. Both uses, say the researchers, offer potential for medical diagnostic applications.

The new holographic imaging technique produces better images than current methods using multiple holograms, and – since it requires fewer measurements and performs computations faster – is easier to implement.

In one study, the researchers produced holograms of Pap smears – which are used to detect signs of cervical cancer – and blood samples, as well as breast tissue samples. The neural network in each case learned to extract and separate the features of the true image of the object from undesired light interference and from other physical by-products of the image reconstruction process.

“These results are broadly applicable to any phase recovery and holographic imaging problem, and this deep-learning–based framework opens up myriad opportunities to design fundamentally new coherent imaging systems, spanning different parts of the electromagnetic spectrum, including visible wavelengths and even X-rays,” says Aydogan Ozcan, an associate director of the UCLA California NanoSystems Institute and the Chancellor’s Professor of Electrical and Computer Engineering at the UCLA Henry Samueli School of Engineering and Applied Science, as well as an HHMI Professor at the Howard Hughes Medical Institute.

According to the researchers, their approach achieved its results without requiring any modeling of light–matter interaction or solving of the wave equation, which can be challenging and time-consuming to model and calculate for each individual sample and form of light.

“This is an exciting achievement since traditional physics-based hologram reconstruction methods have been replaced by a deep-learning–based computational approach,” says Yair Rivenson, postdoctoral scholar at UCLA’s electrical and computer engineering department.

In a second study, the researchers used the approach to improve the resolution and quality of optical microscopic images. This, say the researchers, could help diagnosticians or pathologists look for very small-scale abnormalities in large blood or tissue samples.

For more, see “Deep learning microscopy” and “Phase recovery and holographic image reconstruction using deep learning in neural networks.”

Related articles:
Algorithm beats radiologists in diagnosing x-rays
AI speeds up precision medicine, says IBM Watson study
Deep learning algorithm beats cardiologists at arrhythmia diagnosis

If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :    eeNews on Google News


Linked Articles