東華大學圖書館 |

Advancing Image Reconstruction and Restoration Through Robust Supervised and Generative Models /

Record Type:	Electronic resources : Monograph/item
Title/Author:	Advancing Image Reconstruction and Restoration Through Robust Supervised and Generative Models // Shijun Liang.
Author:	Liang, Shijun,
Description:	1 electronic resource (223 pages)
Notes:	Source: Dissertations Abstracts International, Volume: 86-08, Section: B.
Contained By:	Dissertations Abstracts International86-08B.
Subject:	Biomedical engineering. -
Online resource:	https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31839648
ISBN:	9798304961639

Advancing Image Reconstruction and Restoration Through Robust Supervised and Generative Models /
Liang, Shijun,

Advancing Image Reconstruction and Restoration Through Robust Supervised and Generative Models /Shijun Liang. - 1 electronic resource (223 pages)

Source: Dissertations Abstracts International, Volume: 86-08, Section: B.

Medical imaging is integral to modern clinical workflows, extensively used in diagnosis, prognosis, and treatment planning for numerous diseases. Magnetic Resonance Imaging (MRI) is particularly valuable because it avoids ionizing radiation and provides excellent soft tissue contrast. However, MRI has limitations, such as prolonged scan times that raise imaging costs and increase susceptibility to artifacts like motion. Furthermore, challenges in medical imaging data acquisition and distribution arise due to patient privacy concerns and strict ethical and legal restrictions. Additionally, medical data is highly heterogeneous, varying across institutions, imaging devices, protocols, and patient demographics. Harmonizing data from diverse sources requires extensive preprocessing, complicating data acquisition.This thesis presents algorithms that can be categorized into two main areas. The first area explores MRI reconstruction using limited or data-free machine learning techniques to bridge gaps in data acquisition. Specifically, for cases with limited training data, we propose the LONDN MRI method, which trains on a small set of adaptively chosen neighboring images that are similar to the target image. For data-free scenarios, we advanced the Deep Image Prior, introducing self-guided DIP, which is a self-regularization method leveraging a denoising regularization based on continuously perturbing inputs with random noise and applying network output smoothing to enhance generalization. Inspired by self-guided DIP, we further improve the method efficiency by building upon an insight into the impact of the DIP network input; we introduce Autoencoding Sequential DIP (aSeqDIP), which incorporates a U-Net architecture whose weights are updated sequentially with the input however simply updated in a feed-forward fashion with autoencoding regularization. Our findings indicate that LONDN MRI outperforms the supervised MoDL model by approximately 0.4 dB for MRI reconstruction from limited measurements. On a more challenging datasets such as the FSE MRI dataset, our method achieves a 0.8 dB improvement. Both the Self-Guided DIP and Autoencoding Sequential DIP (aSeqDIP) outperform the state of Art generative model Score MRI model by approximately 0.45 and 0.76 dB, respectively.Secondly, we explore means to improve the reconstruction network's generalization capabilities. We introduce an unrolling method (SMUG) combined with randomized smoothing to counteract effects of worst-case and other perturbations. This approach combines model-based deep unrolling with randomized smoothing, helping to mitigate worst-case perturbations as well as variations such as sampling pattern shifts, differing acceleration factors, and Gaussian noise. Furthermore, a pre-learned diffusion model can act as an effective purifier prior to an unrolled network by incrementally adding Gaussian noise and subsequently removing it, thus serving as a robust noise-removal method for image purification. We implemented the developed method Diffusion Purification in various experiments, particularly on biomedical lesion data, and found it outperforms common robustness approaches, such as adversarial training, randomized smoothing, and baseline methods without robustness enhancements. However, the primary drawback of diffusion models lies in their slow image generation and denoising processes, posing a significant challenge to balancing processing speed and output quality. To handle this issue. we proposed SITCOM that exploits three conditions for achieving measurement-consistent diffusion trajectories with expanded DDIM (DDIM). Building on these conditions, we propose a new optimization-based diffusion reverse sampling method that not only enforces the standard data manifold measurement consistency and forward diffusion consistency, as seen in previous studies, but also incorporates backward diffusion consistency that maintains a diffusion trajectory by optimizing over the input of the pre-trained model at every sampling step. It outperforms the state-of-the-art method DAPS for most of the tasks by 1.2 dB in terms of image quality.

English

ISBN: 9798304961639Subjects--Topical Terms:

535387
Biomedical engineering.
Subjects--Index Terms:

Compress sensing

Advancing Image Reconstruction and Restoration Through Robust Supervised and Generative Models /
LDR:05616nmm a22004213i 4500 001 2434802
005 20251001062528.5
006 m o d
007 cr|nu||||||||
008 261011s2025 miu||||||m |||||||eng d
020 $a 9798304961639
035 $a (MiAaPQD)AAI31839648
035 $a AAI31839648
040 $a MiAaPQD $b eng $c MiAaPQD $e rda
100 1 $a Liang, Shijun, $e author. $0 (orcid)0000-0002-0825-7068 $3 3824909
245 1 0 $a Advancing Image Reconstruction and Restoration Through Robust Supervised and Generative Models / $c Shijun Liang.
264 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2025
300 $a 1 electronic resource (223 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertations Abstracts International, Volume: 86-08, Section: B.
500 $a Advisors: Ravishankar, Saiprasad Committee members: Wang, Rongrong; Alessio, Adam; Bhattacharya, Sudin.
502 $b Ph.D. $c Michigan State University $d 2025.
520 $a Medical imaging is integral to modern clinical workflows, extensively used in diagnosis, prognosis, and treatment planning for numerous diseases. Magnetic Resonance Imaging (MRI) is particularly valuable because it avoids ionizing radiation and provides excellent soft tissue contrast. However, MRI has limitations, such as prolonged scan times that raise imaging costs and increase susceptibility to artifacts like motion. Furthermore, challenges in medical imaging data acquisition and distribution arise due to patient privacy concerns and strict ethical and legal restrictions. Additionally, medical data is highly heterogeneous, varying across institutions, imaging devices, protocols, and patient demographics. Harmonizing data from diverse sources requires extensive preprocessing, complicating data acquisition.This thesis presents algorithms that can be categorized into two main areas. The first area explores MRI reconstruction using limited or data-free machine learning techniques to bridge gaps in data acquisition. Specifically, for cases with limited training data, we propose the LONDN MRI method, which trains on a small set of adaptively chosen neighboring images that are similar to the target image. For data-free scenarios, we advanced the Deep Image Prior, introducing self-guided DIP, which is a self-regularization method leveraging a denoising regularization based on continuously perturbing inputs with random noise and applying network output smoothing to enhance generalization. Inspired by self-guided DIP, we further improve the method efficiency by building upon an insight into the impact of the DIP network input; we introduce Autoencoding Sequential DIP (aSeqDIP), which incorporates a U-Net architecture whose weights are updated sequentially with the input however simply updated in a feed-forward fashion with autoencoding regularization. Our findings indicate that LONDN MRI outperforms the supervised MoDL model by approximately 0.4 dB for MRI reconstruction from limited measurements. On a more challenging datasets such as the FSE MRI dataset, our method achieves a 0.8 dB improvement. Both the Self-Guided DIP and Autoencoding Sequential DIP (aSeqDIP) outperform the state of Art generative model Score MRI model by approximately 0.45 and 0.76 dB, respectively.Secondly, we explore means to improve the reconstruction network's generalization capabilities. We introduce an unrolling method (SMUG) combined with randomized smoothing to counteract effects of worst-case and other perturbations. This approach combines model-based deep unrolling with randomized smoothing, helping to mitigate worst-case perturbations as well as variations such as sampling pattern shifts, differing acceleration factors, and Gaussian noise. Furthermore, a pre-learned diffusion model can act as an effective purifier prior to an unrolled network by incrementally adding Gaussian noise and subsequently removing it, thus serving as a robust noise-removal method for image purification. We implemented the developed method Diffusion Purification in various experiments, particularly on biomedical lesion data, and found it outperforms common robustness approaches, such as adversarial training, randomized smoothing, and baseline methods without robustness enhancements. However, the primary drawback of diffusion models lies in their slow image generation and denoising processes, posing a significant challenge to balancing processing speed and output quality. To handle this issue. we proposed SITCOM that exploits three conditions for achieving measurement-consistent diffusion trajectories with expanded DDIM (DDIM). Building on these conditions, we propose a new optimization-based diffusion reverse sampling method that not only enforces the standard data manifold measurement consistency and forward diffusion consistency, as seen in previous studies, but also incorporates backward diffusion consistency that maintains a diffusion trajectory by optimizing over the input of the pre-trained model at every sampling step. It outperforms the state-of-the-art method DAPS for most of the tasks by 1.2 dB in terms of image quality.
546 $a English
590 $a School code: 0128
650 4 $a Biomedical engineering. $3 535387
650 4 $a Medical imaging. $3 3172799
650 4 $a Bioengineering. $3 657580
653 $a Compress sensing
653 $a Machine learning
653 $a Magnetic Resonance Imaging
653 $a MRI reconstruction
690 $a 0541
690 $a 0202
690 $a 0574
710 2 $a Michigan State University. $b Biomedical Engineering - Doctor of Philosophy. $e degree granting institution. $3 3824910
720 1 $a Ravishankar, Saiprasad $e degree supervisor.
773 0 $t Dissertations Abstracts International $g 86-08B.
790 $a 0128
791 $a Ph.D.
792 $a 2025
856 4 0 $u https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31839648