3D Tensor Based Nonlocal Low Rank Approximation in Dynamic PET Reconstruction

Xie, Nuobei; Chen, Yunmei; Liu, Huafeng

doi:10.3390/s19235299

Cited by 9 publications

(4 citation statements)

References 52 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, tensor decomposition techniques can also be used to denoise an input tensor by finding a low-rank approximation of it, without explicitly rejecting a portion of the extracted components. For example, [55] proposes a method to jointly reconstruct and denoise PET images via low-rank approximating a PET feature tensor X using tensor nuclear norm regularization:…”

Section: Denoising Artifact Removal and Dimensionality Reductionmentioning

confidence: 99%

Tensors for neuroimaging

Erol

Hunyadi

2022

Tensors for Data Processing

View full text Add to dashboard Cite

Section: Denoising Artifact Removal and Dimensionality Reductionmentioning

confidence: 99%

Tensors for neuroimaging

Erol

Hunyadi

2022

Tensors for Data Processing

View full text Add to dashboard Cite

“…In this model, the operational layers consist of : 1) 3 × 3 × 3 convolutional layer; 2) batch normalization (BN) layer; 3) Relu layer; 4) 3 × 3 × 3 stride-2 convolutional layer as the down sampling layer and 5) bilinear interpolation layer as the upsampling layer [17]. Besides, instead of using the concatenation operator, copy and add were Given the fact that no ground truth is available in the real data study, the contrast to noise ratio (CNR) [21] was adopted as the quantitative evaluation:…”

Section: Data Preprocessing and Experimental Implementationmentioning

confidence: 99%

Clinically Translatable Direct Patlak Reconstruction from Dynamic PET with Motion Correction Using Convolutional Neural Network

Xie

Gong

Guo

et al. 2020

Medical Image Computing and Computer Assisted Intervention – MICCAI 2020

View full text Add to dashboard Cite

Patlak model is widely used in 18 F-FDG dynamic positron emission tomography (PET) imaging, where the estimated parametric images reveal important biochemical and physiology information. Because of better noise modeling and more information extracted from raw sinogram, direct Patlak reconstruction gains its popularity over the indirect approach which utilizes reconstructed dynamic PET images alone. As the prerequisite of direct Patlak methods, raw data from dynamic PET are rarely stored in clinics and difficult to obtain. In addition, the direct reconstruction is time-consuming due to the bottleneck of multiple-frame reconstruction. All of these impede the clinical adoption of direct Patlak reconstruction. In this work, we proposed a data-driven framework which maps the dynamic PET images to the high-quality motioncorrected direct Patlak images through a convolutional neural network. For the patient's motion during the long period of dynamic PET scan, we combined the correction with the backward/forward projection in direct reconstruction to better fit the statistical model. Results based on fifteen clinical 18 F-FDG dynamic brain PET datasets demonstrates the superiority of the proposed framework over Gaussian, nonlocal mean and BM4D denoising, regarding the image bias and contrast-to-noise ratio.

show abstract

“…Our work is inspired by the well-known non-local means approach [25], which has been widely used for image denoising [25][26][27][28][29] and restoration/reconstruction [30][31][32][33][34] by exploiting the measure of similarity between the image patches. The nonlocal means approach is based on the idea that in an image, pixels that are similar to each other tend to have similar values.…”

Section: Introductionmentioning

confidence: 99%

Similarity-Driven Fine-Tuning Methods for Regularization Parameter Optimization in PET Image Reconstruction

Zhu

Lee

2023

Sensors

View full text Add to dashboard Cite

We present an adaptive method for fine-tuning hyperparameters in edge-preserving regularization for PET image reconstruction. For edge-preserving regularization, in addition to the smoothing parameter that balances data fidelity and regularization, one or more control parameters are typically incorporated to adjust the sensitivity of edge preservation by modifying the shape of the penalty function. Although there have been efforts to develop automated methods for tuning the hyperparameters in regularized PET reconstruction, the majority of these methods primarily focus on the smoothing parameter. However, it is challenging to obtain high-quality images without appropriately selecting the control parameters that adjust the edge preservation sensitivity. In this work, we propose a method to precisely tune the hyperparameters, which are initially set with a fixed value for the entire image, either manually or using an automated approach. Our core strategy involves adaptively adjusting the control parameter at each pixel, taking into account the degree of patch similarities calculated from the previous iteration within the pixel’s neighborhood that is being updated. This approach allows our new method to integrate with a wide range of existing parameter-tuning techniques for edge-preserving regularization. Experimental results demonstrate that our proposed method effectively enhances the overall reconstruction accuracy across multiple image quality metrics, including peak signal-to-noise ratio, structural similarity, visual information fidelity, mean absolute error, root-mean-square error, and mean percentage error.

show abstract

3D Tensor Based Nonlocal Low Rank Approximation in Dynamic PET Reconstruction

Cited by 9 publications

References 52 publications

Tensors for neuroimaging

Tensors for neuroimaging

Clinically Translatable Direct Patlak Reconstruction from Dynamic PET with Motion Correction Using Convolutional Neural Network

Similarity-Driven Fine-Tuning Methods for Regularization Parameter Optimization in PET Image Reconstruction

Contact Info

Product

Resources

About