De-noising SPECT/PET Images Using Cross-Scale Regularization

Jin, Yinpeng; Angelini, Elsa D.; Esser, Peter D.; Laine, Andrew F.

doi:10.1007/978-3-540-39903-2_5

Cited by 8 publications

(9 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Indeed detailed levels of wavelet sub-band images display high-energy coefficients for both edge features and noise components. A cross-scale regularization scheme was used to guide the threshold level selection from one expansion level to the previous as detailed in [1]. This regularization scheme, which brings spatial adaptivity to the thresholding operator, was motivated by the idea that meaningful signal features show persistent higher levels of coherence across wavelet expansion scales when compared to random noise.…”

Section: Denoising Via Thresholdingmentioning

confidence: 99%

“…In the case of high noise levels, detailed scales of sub-band images are usually dominated by noisy components that are not well handled using traditional thresholding schemes. To address this issue, a cross-scale regularization scheme was introduced in [1], which takes into account across scale coherence of structured signals in wavelet coefficient subbands. Preliminary results showed promising performance in denoising clinical SPECT and PET images for liver and brain studies.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Fusion of brushlet and wavelet denoising methods for nuclear images

Angelini

Esser

Heertum

et al.

2004 2nd IEEE International Symposium on Biomedical Imaging: Macro to Nano (IEEE Cat No. 04EX821)

View full text Add to dashboard Cite

This paper presents preliminary results on the fusion of denoised PET and SPECT data volumes from brushlet and wavelet thresholding methods. Texture-based brushlet denoising is well suited for enhancement of physiological information while wavelet-based denoising is better suited for enhancement of anatomical contours. A three-dimensional multi-scale edge-based data fusion algorithm is applied to combine enhanced data from these two independent denoising methods. Preliminary results with qualitative evaluation of PET and SPECT data by an expert clinician showed great potential of this approach to combine enhancement of both anatomical and physiological signal information for improved image quality.

show abstract

Section: Denoising Via Thresholdingmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Fusion of brushlet and wavelet denoising methods for nuclear images

Angelini

Esser

Heertum

et al.

2004 2nd IEEE International Symposium on Biomedical Imaging: Macro to Nano (IEEE Cat No. 04EX821)

View full text Add to dashboard Cite

show abstract

“…Therefore a more sophisticated "thresholding" scheme was applied (e.g. cross-scale regularization [14]). 2.…”

Section: B Multi-scale Adaptive Thresholdingmentioning

confidence: 99%

“…To recover signal related features in noise dominated wavelet sub-bands, a cross-scale regularization was suggested [14]. First, an edge indication map was constructed using the next higher level of wavelet subbands.…”

Section: Cross-scale Regularizationmentioning

confidence: 99%

Regularization of PET reconstruction using multi-scale adaptive thresholding

Jin

Esser²,

Aikawa³

et al.

The 26th Annual International Conference of the IEEE Engineering in Medicine and Biology Society

View full text Add to dashboard Cite

A multi-scale adaptive thresholding scheme is presented in this study. It was evaluated as a regularization process to filtered back-projection (FBP) for reconstructing clinical PET brain data. Adaptive selection of thresholding operators for each multi-scale sub-band enabled a unified process for noise removal and feature enhancement. A cross-scale regularization process was utilized as an effective signal recovering operator. Together with non-linear thresholding and enhancement operators, they offered remarkable postprocessing to FBP reconstructed data. In addition, such effectiveness was formulated as a regularization process to optimize FBP reconstruction. A comparison study with multiscale regularized FBP (MFBP), standard FBP with clinical settings and iterative reconstruction (OSEM) was reported. The proposed regularization process has shown competitive improvement in the image quality of PET reconstructions when compared to the current state-of-the-art method used in clinical commercial systems (OSEM).

show abstract

“…Jw[k) estimates (and the resulting adaptive thresholds based on these) are determined using denoised coefficients, which axe found to be more reliable.Most recent work by Jin et al uses a refinement called cross-scale regularization (CSR) which increases noise suppression, particularly at the finest scale of decom position[26,27]. Typically, these subbands contain overwhelming am ounts of noise (indeed, this is why these subbands are used for noise level estim ation).…”

mentioning

confidence: 99%

Wavelet-based denoising of cardiac PET data

Green¹

View full text Add to dashboard Cite

A bstractThis thesis focuses on denoising of positron emission tom ography (PET) data.Cardiac P E T scans generated using a rubidium-82 radiotracer are a convenient, n o ninvasive m ethod of diagnosing heart disease, b u t suffer from a high degree of noise.Denoising methods based on the wavelet transform are capable of outperform ing existing clinical m ethods due to their ability to b etter preserve detail while simul taneously suppressing noise a t multiple scales. We investigate th e applicability of recently developed wavelet denoising m ethods to cardiac P E T data. A comprehen sive set of experiments is performed, in which combinations of these techniques are applied to the different decomposition levels of wavelet coefficients. By doing so, we determine th e relevant im portance of each (and the domain in which it is applied)to the overall quality of the denoised result. W ith this inform ation, we propose P E T denoising protocols th a t substantially improve image quality (for static studies) and lead to b etter measures of myocardial perfusion (for dynam ic studies).

show abstract

De-noising SPECT/PET Images Using Cross-Scale Regularization

Cited by 8 publications

References 12 publications

Fusion of brushlet and wavelet denoising methods for nuclear images

Fusion of brushlet and wavelet denoising methods for nuclear images

Regularization of PET reconstruction using multi-scale adaptive thresholding

Wavelet-based denoising of cardiac PET data

Contact Info

Product

Resources

About