Deep image prior for undersampling high-speed photoacoustic microscopy.

dc.contributor.author

Vu, Tri

dc.contributor.author

DiSpirito, Anthony

dc.contributor.author

Li, Daiwei

dc.contributor.author

Wang, Zixuan

dc.contributor.author

Zhu, Xiaoyi

dc.contributor.author

Chen, Maomao

dc.contributor.author

Jiang, Laiming

dc.contributor.author

Zhang, Dong

dc.contributor.author

Luo, Jianwen

dc.contributor.author

Zhang, Yu Shrike

dc.contributor.author

Zhou, Qifa

dc.contributor.author

Horstmeyer, Roarke

dc.contributor.author

Yao, Junjie

dc.date.accessioned

2024-05-14T19:09:21Z

dc.date.available

2024-05-14T19:09:21Z

dc.date.issued

2021-06

dc.description.abstract

Photoacoustic microscopy (PAM) is an emerging imaging method combining light and sound. However, limited by the laser's repetition rate, state-of-the-art high-speed PAM technology often sacrifices spatial sampling density (i.e., undersampling) for increased imaging speed over a large field-of-view. Deep learning (DL) methods have recently been used to improve sparsely sampled PAM images; however, these methods often require time-consuming pre-training and large training dataset with ground truth. Here, we propose the use of deep image prior (DIP) to improve the image quality of undersampled PAM images. Unlike other DL approaches, DIP requires neither pre-training nor fully-sampled ground truth, enabling its flexible and fast implementation on various imaging targets. Our results have demonstrated substantial improvement in PAM images with as few as 1.4 % of the fully sampled pixels on high-speed PAM. Our approach outperforms interpolation, is competitive with pre-trained supervised DL method, and is readily translated to other high-speed, undersampling imaging modalities.

dc.identifier

S2213-5979(21)00027-6

dc.identifier.issn

2213-5979

dc.identifier.issn

2213-5979

dc.identifier.uri

https://hdl.handle.net/10161/30709

dc.language

eng

dc.relation.ispartof

Photoacoustics

dc.relation.isversionof

10.1016/j.pacs.2021.100266

dc.rights.uri

https://creativecommons.org/licenses/by-nc/4.0

dc.subject

Convolutional neural network

dc.subject

Deep image prior

dc.subject

Deep learning

dc.subject

High-speed imaging

dc.subject

Photoacoustic microscopy

dc.subject

Raster scanning

dc.subject

Undersampling

dc.title

Deep image prior for undersampling high-speed photoacoustic microscopy.

dc.type

Journal article

duke.contributor.orcid

Horstmeyer, Roarke|0000-0002-2480-9141

duke.contributor.orcid

Yao, Junjie|0000-0002-2381-706X

pubs.begin-page

100266

pubs.organisational-group

Duke

pubs.organisational-group

Pratt School of Engineering

pubs.organisational-group

School of Medicine

pubs.organisational-group

Clinical Science Departments

pubs.organisational-group

Institutes and Centers

pubs.organisational-group

Biomedical Engineering

pubs.organisational-group

Duke Cancer Institute

pubs.organisational-group

University Institutes and Centers

pubs.organisational-group

Duke Global Health Institute

pubs.organisational-group

Duke Institute for Brain Sciences

pubs.organisational-group

Neurology

pubs.publication-status

Published

pubs.volume

22

Files

Original bundle

Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
Deep image prior for undersampling high-speed photoacoustic microscopy.pdf
Size:
12.21 MB
Format:
Adobe Portable Document Format