Real-Time Cerebral Vessel Segmentation in Laser Speckle Contrast Image Based on Unsupervised Domain Adaptation

Laser speckle contrast imaging (LSCI) is a full-field, high spatiotemporal resolution and low-cost optical technique for measuring blood flow, which has been successfully used for neurovascular imaging. However, due to the low signal–noise ratio and the relatively small sizes, segmenting the cerebra...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Heping Chen, Yan Shi, Bin Bo, Denghui Zhao, Peng Miao, Shanbao Tong, Chunliang Wang
Formato:	article
Lenguaje:	EN
Publicado:	Frontiers Media S.A. 2021
Materias:	laser speckle contrast imaging vessel segmentation CycleGAN domain adaptation blood flow imaging Neurosciences. Biological psychiatry. Neuropsychiatry RC321-571
Acceso en línea:	https://doaj.org/article/d8d65526abb748129c9daf6c1f7e4eba
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:d8d65526abb748129c9daf6c1f7e4eba
record_format	dspace
spelling	oai:doaj.org-article:d8d65526abb748129c9daf6c1f7e4eba2021-12-01T15:57:47ZReal-Time Cerebral Vessel Segmentation in Laser Speckle Contrast Image Based on Unsupervised Domain Adaptation1662-453X10.3389/fnins.2021.755198https://doaj.org/article/d8d65526abb748129c9daf6c1f7e4eba2021-11-01T00:00:00Zhttps://www.frontiersin.org/articles/10.3389/fnins.2021.755198/fullhttps://doaj.org/toc/1662-453XLaser speckle contrast imaging (LSCI) is a full-field, high spatiotemporal resolution and low-cost optical technique for measuring blood flow, which has been successfully used for neurovascular imaging. However, due to the low signal–noise ratio and the relatively small sizes, segmenting the cerebral vessels in LSCI has always been a technical challenge. Recently, deep learning has shown its advantages in vascular segmentation. Nonetheless, ground truth by manual labeling is usually required for training the network, which makes it difficult to implement in practice. In this manuscript, we proposed a deep learning-based method for real-time cerebral vessel segmentation of LSCI without ground truth labels, which could be further integrated into intraoperative blood vessel imaging system. Synthetic LSCI images were obtained with a synthesis network from LSCI images and public labeled dataset of Digital Retinal Images for Vessel Extraction, which were then used to train the segmentation network. Using matching strategies to reduce the size discrepancy between retinal images and laser speckle contrast images, we could further significantly improve image synthesis and segmentation performance. In the testing LSCI images of rodent cerebral vessels, the proposed method resulted in a dice similarity coefficient of over 75%.Heping ChenHeping ChenYan ShiBin BoDenghui ZhaoPeng MiaoShanbao TongChunliang WangFrontiers Media S.A.articlelaser speckle contrast imagingvessel segmentationCycleGANdomain adaptationblood flow imagingNeurosciences. Biological psychiatry. NeuropsychiatryRC321-571ENFrontiers in Neuroscience, Vol 15 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	laser speckle contrast imaging vessel segmentation CycleGAN domain adaptation blood flow imaging Neurosciences. Biological psychiatry. Neuropsychiatry RC321-571
spellingShingle	laser speckle contrast imaging vessel segmentation CycleGAN domain adaptation blood flow imaging Neurosciences. Biological psychiatry. Neuropsychiatry RC321-571 Heping Chen Heping Chen Yan Shi Bin Bo Denghui Zhao Peng Miao Shanbao Tong Chunliang Wang Real-Time Cerebral Vessel Segmentation in Laser Speckle Contrast Image Based on Unsupervised Domain Adaptation
description	Laser speckle contrast imaging (LSCI) is a full-field, high spatiotemporal resolution and low-cost optical technique for measuring blood flow, which has been successfully used for neurovascular imaging. However, due to the low signal–noise ratio and the relatively small sizes, segmenting the cerebral vessels in LSCI has always been a technical challenge. Recently, deep learning has shown its advantages in vascular segmentation. Nonetheless, ground truth by manual labeling is usually required for training the network, which makes it difficult to implement in practice. In this manuscript, we proposed a deep learning-based method for real-time cerebral vessel segmentation of LSCI without ground truth labels, which could be further integrated into intraoperative blood vessel imaging system. Synthetic LSCI images were obtained with a synthesis network from LSCI images and public labeled dataset of Digital Retinal Images for Vessel Extraction, which were then used to train the segmentation network. Using matching strategies to reduce the size discrepancy between retinal images and laser speckle contrast images, we could further significantly improve image synthesis and segmentation performance. In the testing LSCI images of rodent cerebral vessels, the proposed method resulted in a dice similarity coefficient of over 75%.
format	article
author	Heping Chen Heping Chen Yan Shi Bin Bo Denghui Zhao Peng Miao Shanbao Tong Chunliang Wang
author_facet	Heping Chen Heping Chen Yan Shi Bin Bo Denghui Zhao Peng Miao Shanbao Tong Chunliang Wang
author_sort	Heping Chen
title	Real-Time Cerebral Vessel Segmentation in Laser Speckle Contrast Image Based on Unsupervised Domain Adaptation
title_short	Real-Time Cerebral Vessel Segmentation in Laser Speckle Contrast Image Based on Unsupervised Domain Adaptation
title_full	Real-Time Cerebral Vessel Segmentation in Laser Speckle Contrast Image Based on Unsupervised Domain Adaptation
title_fullStr	Real-Time Cerebral Vessel Segmentation in Laser Speckle Contrast Image Based on Unsupervised Domain Adaptation
title_full_unstemmed	Real-Time Cerebral Vessel Segmentation in Laser Speckle Contrast Image Based on Unsupervised Domain Adaptation
title_sort	real-time cerebral vessel segmentation in laser speckle contrast image based on unsupervised domain adaptation
publisher	Frontiers Media S.A.
publishDate	2021
url	https://doaj.org/article/d8d65526abb748129c9daf6c1f7e4eba
work_keys_str_mv	AT hepingchen realtimecerebralvesselsegmentationinlaserspecklecontrastimagebasedonunsuperviseddomainadaptation AT hepingchen realtimecerebralvesselsegmentationinlaserspecklecontrastimagebasedonunsuperviseddomainadaptation AT yanshi realtimecerebralvesselsegmentationinlaserspecklecontrastimagebasedonunsuperviseddomainadaptation AT binbo realtimecerebralvesselsegmentationinlaserspecklecontrastimagebasedonunsuperviseddomainadaptation AT denghuizhao realtimecerebralvesselsegmentationinlaserspecklecontrastimagebasedonunsuperviseddomainadaptation AT pengmiao realtimecerebralvesselsegmentationinlaserspecklecontrastimagebasedonunsuperviseddomainadaptation AT shanbaotong realtimecerebralvesselsegmentationinlaserspecklecontrastimagebasedonunsuperviseddomainadaptation AT chunliangwang realtimecerebralvesselsegmentationinlaserspecklecontrastimagebasedonunsuperviseddomainadaptation
_version_	1718404843653038080

Real-Time Cerebral Vessel Segmentation in Laser Speckle Contrast Image Based on Unsupervised Domain Adaptation

Ejemplares similares