Deep learning has been experiencing a true renaissance
especially over the last decade, and it uses multi-layered artificial neural
networks for automated analysis of data. Deep learning is one of the most
exciting forms of machine learning that is behind several recent leapfrog
advances in technology including for example real-time speech recognition and translation
as well image/video labeling and captioning, among many others. Especially in
image analysis, deep learning shows significant promise for automated search
and labeling of features of interest, such as abnormal regions in a medical
image.
Now, UCLA researchers have demonstrated a new use for deep
learning – this time to reconstruct a hologram and form a microscopic image of
an object. In a recent article that is published in Light: Science & Applications, a journal of the Springer Nature, UCLA researchers
have demonstrated that a neural network can learn to perform phase recovery and
holographic image reconstruction after appropriate training. This deep
learning-based approach provides a fundamentally new framework to conduct
holographic imaging and compared to existing approaches it is significantly
faster to compute and reconstructs improved images of the objects using a
single hologram, such that it requires fewer measurements in addition to being
computationally faster.
This research was led by Dr. 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, along with Dr. Yair Rivenson, a postdoctoral
scholar, and Yibo Zhang, a graduate student, both at the UCLA electrical and
computer engineering department.
The authors validated this deep learning based approach by
reconstructing holograms of various samples including blood and Pap smears
(used for screening of cervical cancer) as well as thin sections of tissue
samples used in pathology, all of which demonstrated successful elimination of spatial
artifacts that arise from the lost phase information at the hologram recording
process. Stated differently, after its training the neural network has learned
to extract and separate the spatial features of the true image of the object
from undesired light interference and related artifacts. Remarkably, this deep
learning based hologram recovery has been achieved without any modeling of
light-matter interaction or a solution of the wave equation. “This is an
exciting achievement since traditional physics-based hologram reconstruction
methods have been replaced by a deep learning based computational approach”
said Rivenson.
“These results are broadly applicable to any phase recovery
and holographic imaging problem, and this deep learning based framework opens
up a myriad of opportunities to design fundamentally new coherent imaging
systems, spanning different parts of the electromagnetic spectrum, including
visible wavelengths as well as the X-ray regime” added Ozcan, who is also an
HHMI Professor with the Howard Hughes Medical Institute.
Other members of the research team were Harun Günaydın and
Da Teng, members of the Ozcan Research Lab at UCLA.
Ozcan’s research is supported by a Presidential Early Career
Award for Scientists and Engineers, the Army Research Office, the National
Science Foundation, the Office of Naval Research, the National Institutes of
Health, the Howard Hughes Medical Institute, the Vodafone Americas Foundation,
the Mary Kay Foundation and the Steven and Alexandra Cohen Foundation.