東華大學圖書館 |

Novel Algorithms for Motion Correction and Image Processing in Positron Emission Tomography.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Novel Algorithms for Motion Correction and Image Processing in Positron Emission Tomography./
作者:	Ren, Silin.
出版者:	Ann Arbor : ProQuest Dissertations & Theses, : 2018,
面頁冊數:	176 p.
附註:	Source: Dissertations Abstracts International, Volume: 80-02, Section: B.
Contained By:	Dissertations Abstracts International80-02B.
標題:	Biomedical engineering. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=10927885
ISBN:	9780438194557

Novel Algorithms for Motion Correction and Image Processing in Positron Emission Tomography.
Ren, Silin.

Novel Algorithms for Motion Correction and Image Processing in Positron Emission Tomography. - Ann Arbor : ProQuest Dissertations & Theses, 2018 - 176 p.

Source: Dissertations Abstracts International, Volume: 80-02, Section: B.

Thesis (Ph.D.)--Yale University, 2018.

This item is not available from ProQuest Dissertations & Theses.

Positron emission tomography (PET) is a non-invasive functional imaging modality that provides molecular level information with a wide range of applications in clinical and research studies. These include disease diagnosis, staging, and treatment planning in oncology, cardiology, and neurology, and understanding physiological processes and determining functional characterization of tissues in research. A PET study starts with the injection of a radiopharmaceutical, followed by a continuous scan of several minutes to hours. Then, PET images are generated by image reconstruction algorithms for further analysis. For the past few decades, there have been great advances in PET instrumentation, followed by image reconstruction algorithms. But there are still many challenges including improving image quality and more sophisticated extraction of information from image data. This dissertation provides novel algorithms to address these two challenges by: proposing 1) novel respiratory motion correction methods to improve image quality, and 2) novel image processing algorithms to better extract information in two ways: automatically delineating organs for PET quantitative analysis, and automatically classifying neurological disorders using PET synaptic density imaging through machine learning classification algorithms. This work is comprised of four projects, as described in the following paragraphs: The first and second projects were aimed at improving image quality and reducing motion-induced blurring by introducing a novel data-driven respiratory motion estimation method: Centroid-of-distribution (COD) and data-driven event-by-event (EBE) respiratory motion correction. Due to the long acquisition time of PET studies, respiratory motion is inevitable and typically leads to image blurring. To reduce motion-induced blurring, respiratory motion correction approaches have been proposed, but all require respiratory motion input, which is typically acquired from external devices. Because of the complexity in system set up and increased expense, external devices are not commonly employed in clinical studies. Therefore, it is desirable to develop data-driven respiratory motion detection methods to make respiratory motion correction feasible for PET studies without external motion tracking devices. In the first project, based on PET listmode data with time-of-flight (TOF) information, a novel data-driven respiratory motion estimation method named Centroid of distribution (COD) was proposed and evaluated by comparing with motion signals from external devices. Based on COD signals, the first data-driven event-by-event respiratory motion correction method was developed and shown to provide comparable motion correction results to the external device-based method. The second project further developed the first data-driven event-by-event nonrigid respiratory motion correction method (DD-EBE-NRMC) addressing the limitation of only correcting for rigid motion in the first project. In this work, COD was compared with PCA (principal component analysis), which is one of the most well-known data-driven motion estimation methods. Both COD-and PCA-derived motion signals were used as input to DD-EBE-NRMC, and showed comparable motion correction results as the external device-based method for datasets from three studies. The third project was aimed at proposing an automatic segmentation method for dynamic PET abdominal studies to address the issues of manual segmentation, which is time consuming and operator dependent. It also aimed to overcome the challenges in abdominal segmentation due to great variations in organ shape and location, and the close proximity among neighboring organs. This work proposed an atlas-based multi-organ segmentation method employing a maximum a posteriori (MAP) framework segmenting pancreas, liver, spleen and kidneys simultaneously. The algorithm was evaluated by comparison with manual segmentations in terms of segmented ROI shape, extracted time activity curves (TACs), and kinetic modeling results, and the proposed automated segmentation method was shown to provide reliable ROIs for quantitative analysis. In the last project, to help with the diagnosis of various neurological disorders, image classification techniques based on PET synaptic density imaging using a novel tracer [11C]UCB-J were investigated. A classical support vector machine (SVM) framework was employed based on features extracted from binding potential (BPND) images derived from kinetic modeling.

ISBN: 9780438194557Subjects--Topical Terms:

535387
Biomedical engineering.

Novel Algorithms for Motion Correction and Image Processing in Positron Emission Tomography.
LDR:05773nmm a2200349 4500 001 2210527
005 20191121124234.5
008 201008s2018 ||||||||||||||||| ||eng d
020 $a 9780438194557
035 $a (MiAaPQ)AAI10927885
035 $a AAI10927885
040 $a MiAaPQ $c MiAaPQ
100 1 $a Ren, Silin. $3 3437666
245 1 0 $a Novel Algorithms for Motion Correction and Image Processing in Positron Emission Tomography.
260 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2018
300 $a 176 p.
500 $a Source: Dissertations Abstracts International, Volume: 80-02, Section: B.
500 $a Publisher info.: Dissertation/Thesis.
500 $a Advisor: Carson, Richard E.
502 $a Thesis (Ph.D.)--Yale University, 2018.
506 $a This item is not available from ProQuest Dissertations & Theses.
506 $a This item must not be added to any third party search indexes.
506 $a This item must not be sold to any third party vendors.
520 $a Positron emission tomography (PET) is a non-invasive functional imaging modality that provides molecular level information with a wide range of applications in clinical and research studies. These include disease diagnosis, staging, and treatment planning in oncology, cardiology, and neurology, and understanding physiological processes and determining functional characterization of tissues in research. A PET study starts with the injection of a radiopharmaceutical, followed by a continuous scan of several minutes to hours. Then, PET images are generated by image reconstruction algorithms for further analysis. For the past few decades, there have been great advances in PET instrumentation, followed by image reconstruction algorithms. But there are still many challenges including improving image quality and more sophisticated extraction of information from image data. This dissertation provides novel algorithms to address these two challenges by: proposing 1) novel respiratory motion correction methods to improve image quality, and 2) novel image processing algorithms to better extract information in two ways: automatically delineating organs for PET quantitative analysis, and automatically classifying neurological disorders using PET synaptic density imaging through machine learning classification algorithms. This work is comprised of four projects, as described in the following paragraphs: The first and second projects were aimed at improving image quality and reducing motion-induced blurring by introducing a novel data-driven respiratory motion estimation method: Centroid-of-distribution (COD) and data-driven event-by-event (EBE) respiratory motion correction. Due to the long acquisition time of PET studies, respiratory motion is inevitable and typically leads to image blurring. To reduce motion-induced blurring, respiratory motion correction approaches have been proposed, but all require respiratory motion input, which is typically acquired from external devices. Because of the complexity in system set up and increased expense, external devices are not commonly employed in clinical studies. Therefore, it is desirable to develop data-driven respiratory motion detection methods to make respiratory motion correction feasible for PET studies without external motion tracking devices. In the first project, based on PET listmode data with time-of-flight (TOF) information, a novel data-driven respiratory motion estimation method named Centroid of distribution (COD) was proposed and evaluated by comparing with motion signals from external devices. Based on COD signals, the first data-driven event-by-event respiratory motion correction method was developed and shown to provide comparable motion correction results to the external device-based method. The second project further developed the first data-driven event-by-event nonrigid respiratory motion correction method (DD-EBE-NRMC) addressing the limitation of only correcting for rigid motion in the first project. In this work, COD was compared with PCA (principal component analysis), which is one of the most well-known data-driven motion estimation methods. Both COD-and PCA-derived motion signals were used as input to DD-EBE-NRMC, and showed comparable motion correction results as the external device-based method for datasets from three studies. The third project was aimed at proposing an automatic segmentation method for dynamic PET abdominal studies to address the issues of manual segmentation, which is time consuming and operator dependent. It also aimed to overcome the challenges in abdominal segmentation due to great variations in organ shape and location, and the close proximity among neighboring organs. This work proposed an atlas-based multi-organ segmentation method employing a maximum a posteriori (MAP) framework segmenting pancreas, liver, spleen and kidneys simultaneously. The algorithm was evaluated by comparison with manual segmentations in terms of segmented ROI shape, extracted time activity curves (TACs), and kinetic modeling results, and the proposed automated segmentation method was shown to provide reliable ROIs for quantitative analysis. In the last project, to help with the diagnosis of various neurological disorders, image classification techniques based on PET synaptic density imaging using a novel tracer [11C]UCB-J were investigated. A classical support vector machine (SVM) framework was employed based on features extracted from binding potential (BPND) images derived from kinetic modeling.
590 $a School code: 0265.
650 4 $a Biomedical engineering. $3 535387
650 4 $a Medical imaging. $3 3172799
650 4 $a Computer science. $3 523869
690 $a 0541
690 $a 0574
690 $a 0984
710 2 $a Yale University. $3 515640
773 0 $t Dissertations Abstracts International $g 80-02B.
790 $a 0265
791 $a Ph.D.
792 $a 2018
793 $a English
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=10927885