FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2553514950> ?p ?o ?g. }

• W2553514950 endingPage "205" @default.
• W2553514950 startingPage "205" @default.
• W2553514950 abstract "205 Objectives The attenuation correction of time-of-flight (TOF) PET/MR scanner can be computed either by maximum likelihood activity and attenuation reconstruction (MLAA) using TOF information, or by MR-based attenuation correction. Although the MLAA algorithm has been theoretically presented and validated, the quality of estimated transmission image is not always acceptable depending on the amount of emission counts in a region. On the other hand, MR-based attenuation correction is known for the large bias in certain tissue and bone. To address above challenging issues, we propose a MLAA algorithm with MR-based anatomical prior to accurately estimation both emission and attenuation image using the TOF information as well as the high resolution MR image. We also compared the performance with the conventional MLAA using only TOF PET data, attenuation image derived from Dixon-based MR images, and the MLAA method using MR-based anatomical prior. Methods The MLAA with MR-prior iterates two steps: one step is the estimation of activity image using TOF-based separable quadratic surrogate (SQS) with a quadratic regularization; and another step is the estimation of attenuation image using the maximum likelihood transmission (MLTR) and the MR-prior based regularization with a patch-based non-local weight. Specifically, the non-local weight calculates the similarity between patches in the MR image, and the corresponding similarity weight is used to estimate the attenuation image. In this framework, we can also use multiple MR images in the prior to facilitate the estimation of attenuation image. For example, we used Dixon-based fat, in-phase, water MR images in our patient study. We exploit an alternating direction method of multipliers (ADMM) algorithm to optimize the cost function with multiple regularization terms. To evaluate the performance of the proposed method, both simulation data with a XCAT phantom and a patient data (male, 56 years old) with lung cancer and bone metastases were used. The proposed method demonstrate superior image quality in simulation, compared with traditional methods. In patient study, the GE SIGNA TOF-PET/MR scanner with 420 ns time resolution is used. We compared the qualities of attenuation images using the conventional MLAA, Dixon-based mapping and the proposed method as well as the emission images after attenuation correction. We also compared the contrast between tumor region and background region in the final emission images of these methods. Results In comparison of attenuation images in figure 1(a), the attenuation image of conventional MLAA is very noisy. It also shows artifacts, such as additional intensity in the air and missing intensity in the chest wall. Dixon-based attenuation image shows a clear result, however, the attenuation values are piecewise constant. In the MLAA with a MR-prior, the image quality is significantly better than the conventional MLAA, and we can see details clearly. In figure 1(b), we compared activity images before and after attenuation correction using the conventional MLAA, Dixon-based mapping and the proposed method. The proposed method shows superior image quality, particularly in the bone region that is important for this patient with bone metastases. More specifically, we confirmed that the tumor-to-muscle ratio of the proposed method is higher than other methods as shown in figure 1(c). Conclusions The MR-prior based MLAA is successfully developed and validated using both simulation and patient study. The results indicated that it can achieve superior quality in both emission and attenuation images because of the utilization of the TOF information as well as the MR information in the estimation." @default.
• W2553514950 created "2016-11-30" @default.
• W2553514950 creator A5038806083 @default.
• W2553514950 creator A5053436378 @default.
• W2553514950 creator A5058429770 @default.
• W2553514950 creator A5075254600 @default.
• W2553514950 creator A5078648367 @default.
• W2553514950 date "2016-05-01" @default.
• W2553514950 modified "2023-09-25" @default.
• W2553514950 title "Accurate Attenuation Correction of TOF-PET/MR Scanner using Both MLAA and Anatomic Prior" @default.
• W2553514950 hasPublicationYear "2016" @default.
• W2553514950 type Work @default.
• W2553514950 sameAs 2553514950 @default.
• W2553514950 citedByCount "0" @default.
• W2553514950 crossrefType "journal-article" @default.
• W2553514950 hasAuthorship W2553514950A5038806083 @default.
• W2553514950 hasAuthorship W2553514950A5053436378 @default.
• W2553514950 hasAuthorship W2553514950A5058429770 @default.
• W2553514950 hasAuthorship W2553514950A5075254600 @default.
• W2553514950 hasAuthorship W2553514950A5078648367 @default.
• W2553514950 hasConcept C115961682 @default.
• W2553514950 hasConcept C120665830 @default.
• W2553514950 hasConcept C121332964 @default.
• W2553514950 hasConcept C123688308 @default.
• W2553514950 hasConcept C153180895 @default.
• W2553514950 hasConcept C154945302 @default.
• W2553514950 hasConcept C184652730 @default.
• W2553514950 hasConcept C2776135515 @default.
• W2553514950 hasConcept C2779751349 @default.
• W2553514950 hasConcept C31972630 @default.
• W2553514950 hasConcept C33923547 @default.
• W2553514950 hasConcept C41008148 @default.
• W2553514950 hasConcept C55020928 @default.
• W2553514950 hasConceptScore W2553514950C115961682 @default.
• W2553514950 hasConceptScore W2553514950C120665830 @default.
• W2553514950 hasConceptScore W2553514950C121332964 @default.
• W2553514950 hasConceptScore W2553514950C123688308 @default.
• W2553514950 hasConceptScore W2553514950C153180895 @default.
• W2553514950 hasConceptScore W2553514950C154945302 @default.
• W2553514950 hasConceptScore W2553514950C184652730 @default.
• W2553514950 hasConceptScore W2553514950C2776135515 @default.
• W2553514950 hasConceptScore W2553514950C2779751349 @default.
• W2553514950 hasConceptScore W2553514950C31972630 @default.
• W2553514950 hasConceptScore W2553514950C33923547 @default.
• W2553514950 hasConceptScore W2553514950C41008148 @default.
• W2553514950 hasConceptScore W2553514950C55020928 @default.
• W2553514950 hasLocation W25535149501 @default.
• W2553514950 hasOpenAccess W2553514950 @default.
• W2553514950 hasPrimaryLocation W25535149501 @default.
• W2553514950 hasRelatedWork W2236883999 @default.
• W2553514950 hasRelatedWork W2294224257 @default.
• W2553514950 hasRelatedWork W2439419579 @default.
• W2553514950 hasRelatedWork W2544579720 @default.
• W2553514950 hasRelatedWork W2558741040 @default.
• W2553514950 hasRelatedWork W272795361 @default.
• W2553514950 hasRelatedWork W2731493393 @default.
• W2553514950 hasRelatedWork W2755563559 @default.
• W2553514950 hasRelatedWork W2766775968 @default.
• W2553514950 hasRelatedWork W2766818712 @default.
• W2553514950 hasRelatedWork W2789642020 @default.
• W2553514950 hasRelatedWork W2798093278 @default.
• W2553514950 hasRelatedWork W280100947 @default.
• W2553514950 hasRelatedWork W2809386964 @default.
• W2553514950 hasRelatedWork W2979766213 @default.
• W2553514950 hasRelatedWork W3047648963 @default.
• W2553514950 hasRelatedWork W3114427144 @default.
• W2553514950 hasRelatedWork W3133336535 @default.
• W2553514950 hasRelatedWork W3202929504 @default.
• W2553514950 hasRelatedWork W350017381 @default.
• W2553514950 hasVolume "57" @default.
• W2553514950 isParatext "false" @default.
• W2553514950 isRetracted "false" @default.
• W2553514950 magId "2553514950" @default.
• W2553514950 workType "article" @default.