東華大學圖書館 |

語系: 繁體中文

說明(常見問題)

回圖書館首頁

手機版館藏查詢

登入

回首頁

切換: 標籤 | MARC模式 | ISBD

Compressive Diffuse Optical Tomograp...

Yao, Ruouang.

FindBook

Google Book

Amazon

博客來

Compressive Diffuse Optical Tomography Based on Structured Light and Monte Carlo Methods.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Compressive Diffuse Optical Tomography Based on Structured Light and Monte Carlo Methods./
作者:	Yao, Ruouang.
出版者:	Ann Arbor : ProQuest Dissertations & Theses, : 2018,
面頁冊數:	132 p.
附註:	Source: Dissertation Abstracts International, Volume: 80-03(E), Section: B.
Contained By:	Dissertation Abstracts International80-03B(E).
標題:	Biomedical engineering. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=10843484
ISBN:	9780438516120

Compressive Diffuse Optical Tomography Based on Structured Light and Monte Carlo Methods.
Yao, Ruouang.

Compressive Diffuse Optical Tomography Based on Structured Light and Monte Carlo Methods. - Ann Arbor : ProQuest Dissertations & Theses, 2018 - 132 p.

Source: Dissertation Abstracts International, Volume: 80-03(E), Section: B.

Thesis (Ph.D.)--Rensselaer Polytechnic Institute, 2018.

Optical tomography with near infrared light (NIR) has gone through a rapid development over the last three decades with numerous applications in clinical and preclinical settings. Diffuse Optical Tomography (DOT) is widely acknowledged to provide irreplaceable functional information for brain imaging, optical mammography, peripheral vasculature diseases, etc. Also, Fluorescence Molecular Tomography (FMT) is becoming more central in preclinical applications, such as imaging small animals or engineered tissues, thanks to the ever-increasing availability of fluorescent molecular probes and its high sensitivity with unique multiplexing potential. However, technical implementations of optical tomography have been relying on raster scanning approaches or fixed punctual sources that are time multiplexed to capture tomographic information. Such configurations are slow, prone to sampling errors (object not in the plane of the optodes), and do not allow efficient scaling for whole body imaging.

ISBN: 9780438516120Subjects--Topical Terms:

535387
Biomedical engineering.

Compressive Diffuse Optical Tomography Based on Structured Light and Monte Carlo Methods.
LDR:05839nmm a2200349 4500 001 2200021
005 20181210125319.5
008 201008s2018 ||||||||||||||||| ||eng d
020 $a 9780438516120
035 $a (MiAaPQ)AAI10843484
035 $a (MiAaPQ)rpi:11382
035 $a AAI10843484
040 $a MiAaPQ $c MiAaPQ
100 1 $a Yao, Ruouang. $3 3426771
245 1 0 $a Compressive Diffuse Optical Tomography Based on Structured Light and Monte Carlo Methods.
260 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2018
300 $a 132 p.
500 $a Source: Dissertation Abstracts International, Volume: 80-03(E), Section: B.
500 $a Adviser: Xavier Intes.
502 $a Thesis (Ph.D.)--Rensselaer Polytechnic Institute, 2018.
520 $a Optical tomography with near infrared light (NIR) has gone through a rapid development over the last three decades with numerous applications in clinical and preclinical settings. Diffuse Optical Tomography (DOT) is widely acknowledged to provide irreplaceable functional information for brain imaging, optical mammography, peripheral vasculature diseases, etc. Also, Fluorescence Molecular Tomography (FMT) is becoming more central in preclinical applications, such as imaging small animals or engineered tissues, thanks to the ever-increasing availability of fluorescent molecular probes and its high sensitivity with unique multiplexing potential. However, technical implementations of optical tomography have been relying on raster scanning approaches or fixed punctual sources that are time multiplexed to capture tomographic information. Such configurations are slow, prone to sampling errors (object not in the plane of the optodes), and do not allow efficient scaling for whole body imaging.
520 $a Over the past decade, the commercial availability of spatial light modulators (SLMs) has led to a surge in new optical imaging techniques based on structured light illumination and camera-based detection strategies. These instrumental developments are poised to transform optical imaging as they enable faster data acquisition, larger imaging field-of-view (FOV), and improved measurement robustness. It has also been demonstrated that DOT based on double light modulator architecture can be implemented at very high speed. We can even couple digital mirror devices (DMDs) with a time-resolved photo-multiplier tube (PMT) and spectrophotometer to acquire information-enriched 4D data cubes.
520 $a However, the introduction of new hardware also brings requirements and challenges for developing dedicated simulation tools and reconstruction methods. For example: 1) tools for modeling light transport with arbitrary extended sources or detection schemes have not been developed, 2) the calculation of sensitivity matrices (Jacobians) for DOT/FMT has been following the old approach and not efficient for wide-field, multi-spectral applications, 3) different structured light illumination and single-pixel detection strategies need systematic evaluation, 4) the inverse problem of DOT/FMT is well known to be ill-posed, ill-conditioned and the situation can be further impaired by the wide-field configurations. In this thesis, we propose new computational tools and formulations to improve the quality and efficiency of wide-field DOT/FMT reconstructions and make this novel technique applicable for both clinical and preclinical research.
520 $a We made efforts from all computational perspectives, including 1) developing more accurate and flexible tools for forward modeling of light transport inside biological tissues, 2) developing more efficient formulations to build Jacobians for wide-field, multi-spectral configurations, 3) comparing different structured light strategies and determining the optimal combination for our own application, 4) improving the property of wide-field Jacobians for the state-of-art inverse solvers to obtain optimal reconstructions. Besides, inspired by the recent success of Deep Learning, especially in image-related areas, we have also investigated using a Convolutional Neural Network as the solver for inverse problems of DOT/FMT.
520 $a The chapters of this thesis are organized as follows: In Chapter 1, we provide a review of the techniques that are critical to understand the materials of other chapters, i.e., the principles of DOT/FMT, Monte Carlo (MC) methods for modeling light propagation, structured light illumination and single-pixel detection, and Compressive Sensing (CS). In Chapter 2, we report a generalized mesh-based MC algorithm to support a variety of wide-field illumination methods including arbitrary 2-D patterns and wide-field detection such as a free-space CCD camera. Chapter 3 introduces an approach "photon replay" to directly compute spatially and temporarily resolved Jacobians for either absorption or scattering changes via perturbation Monte Carlo (pMC). In Chapter 4 we use the newly developed wide-field MC tools to perform a simulation study that compares different structured illumination and single-pixel detection strategies. A CS-based preconditioning method is then presented to improve the property of wide-field FMT Jacobians and reconstruction results. In Chapter 5, we propose a deep Convolutional Neural Network (CNN) named Net-FLICS, to directly reconstruct both intensity and lifetime images in 2D Compressive Macroscopic Fluorescence Lifetime Imaging (MFLI). Chapter 6 summarizes all the efforts and achievements in this thesis. Future directions that's worth investigating and relevant to the topics in this research are also discussed.
590 $a School code: 0185.
650 4 $a Biomedical engineering. $3 535387
650 4 $a Medical imaging. $3 3172799
690 $a 0541
690 $a 0574
710 2 $a Rensselaer Polytechnic Institute. $b Biomedical Engineering. $3 3178857
773 0 $t Dissertation Abstracts International $g 80-03B(E).
790 $a 0185
791 $a Ph.D.
792 $a 2018
793 $a English
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=10843484