자료유형  

 학위논문

Control Number  

0017164116

International Standard Book Number  

9798384069218

Dewey Decimal Classification Number  

616

Main Entry-Personal Name  

Buelo, Collin.

Publication, Distribution, etc. (Imprint  

[S.l.] : The University of Wisconsin - Madison., 2024

Publication, Distribution, etc. (Imprint  

Ann Arbor : ProQuest Dissertations & Theses, 2024

Physical Description  

99 p.

General Note  

Source: Dissertations Abstracts International, Volume: 86-03, Section: B.

General Note  

Advisor: Hernando, Diego.

Dissertation Note  

Thesis (Ph.D.)--The University of Wisconsin - Madison, 2024.

Summary, Etc.  

요약Liver iron overload is a frequent complication of certain genetic disorders, as well as a consequence of repeated blood transfusions as treatment for a variety of conditions including sickle cell anemia and side effect of treatments for certain cancers. Quantitative Susceptibility Mapping (QSM) is a promising method for assessing liver iron concentration (LIC) through the measurement of magnetic susceptibility, a fundamental tissue property. Susceptibility as a measurement of iron is unconfounded by properties that may confound other MRI-based non-invasive iron measurement methods. R2 may be confounded by the presence of fat and the microscopic distribution of iron, while R2* may also be confounded by the microscopic distribution of iron. Therefore, QSM may enable widespread non-invasive, unconfounded LIC measurement in patients with or at risk of developing iron overload.In this thesis, a previously proposed QSM method is assessed in a study across multiple centers and MR vendors at two field strengths. Reproducibility across center and field strength is examined, as well as test-retest repeatability. This study is a comprehensive analysis of the repeatability and reproducibility of QSM. Furthermore, a deep learning-based QSM method is proposed to overcome the limitations of existing QSM methods, primarily that these existing methods have been shown to underestimate susceptibility compared to ground truth. The performance of this deep learning QSM method is compared with the previously evaluated liver QSM method across 4 centers at a single field strength. The deep learning QSM method is shown to estimate higher susceptibility than the previous method, which may advance the establishment of accurate QSM that addresses prior underestimation challenges. Finally, the performance of the previously proposed QSM method and the proposed deep learning-based QSM method is analyzed both in digital phantoms and in a multicenter study across several imaging resolutions. The previously proposed QSM method is shown to have consistent performance across resolution. The performance of the deep learning-based method improves at higher acquisition resolution, with lower variability in estimated susceptibility. Overall, this work shows improvement in QSM measurement with the proposed deep learning-based method with higher, nearly isotropic spatial resolution.This thesis advances QSM as a quantitative biomarker of liver iron, both through validation of existing methods, and through development and evaluation a of novel deep learning-based method that has the potential to overcome the limitations of existing methods.

Subject Added Entry-Topical Term  

Medical imaging.

Subject Added Entry-Topical Term  

Bioinformatics.

Subject Added Entry-Topical Term  

Biophysics.

Index Term-Uncontrolled  

Liver iron

Index Term-Uncontrolled  

Quantitative Susceptibility Mapping

Index Term-Uncontrolled  

Liver iron concentration

Index Term-Uncontrolled  

Higher susceptibility

Index Term-Uncontrolled  

Deep learning

Added Entry-Corporate Name  

The University of Wisconsin - Madison Medical Physics

Host Item Entry  

Dissertations Abstracts International. 86-03B.

Electronic Location and Access  

로그인을 한후 보실 수 있는 자료입니다.

Control Number  

joongbu:656317