An Effective Post-Filtering Framework for 3-D PET Image Denoising Based on Noise and Sensitivity Characteristics
Positron emission tomography (PET) images usually suffer from a noticeable amount of statistical noise. In order to reduce this noise, a post-filtering process is usually adopted. However, the performance of this approach is limited because the denoising process is mostly performed on the basis of t...
Ausführliche Beschreibung
Gespeichert in:
| Autor*in: |
Ji Hye Kim [verfasserIn] |
|---|
| Format: |
Artikel |
|---|---|
| Sprache: |
Englisch |
| Erschienen: |
2015 |
|---|
| Übergeordnetes Werk: |
Enthalten in: IEEE transactions on nuclear science - New York, NY : IEEE, 1963, 62(2015), 1, Seite 137-147 |
|---|---|
| Übergeordnetes Werk: |
volume:62 ; year:2015 ; number:1 ; pages:137-147 |
| Links: |
|---|
| DOI / URN: |
10.1109/TNS.2014.2360176 |
|---|
| Katalog-ID: |
OLC196622365X |
|---|
| LEADER | 01000caa a2200265 4500 | ||
|---|---|---|---|
| 001 | OLC196622365X | ||
| 003 | DE-627 | ||
| 005 | 20230525173913.0 | ||
| 007 | tu | ||
| 008 | 160206s2015 xx ||||| 00| ||eng c | ||
| 024 | 7 | |a 10.1109/TNS.2014.2360176 |2 doi | |
| 028 | 5 | 2 | |a PQ20160617 |
| 035 | |a (DE-627)OLC196622365X | ||
| 035 | |a (DE-599)GBVOLC196622365X | ||
| 035 | |a (PRQ)c1776-ac9a27912cb4c6ff7cd557cb2c8263ed56b68ec02f1e24b8d28d5d59e2b9f7880 | ||
| 035 | |a (KEY)0054996720150000062000100137effectivepostfilteringframeworkfor3dpetimagedenois | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rakwb | ||
| 041 | |a eng | ||
| 082 | 0 | 4 | |a 620 |q DNB |
| 100 | 0 | |a Ji Hye Kim |e verfasserin |4 aut | |
| 245 | 1 | 3 | |a An Effective Post-Filtering Framework for 3-D PET Image Denoising Based on Noise and Sensitivity Characteristics |
| 264 | 1 | |c 2015 | |
| 336 | |a Text |b txt |2 rdacontent | ||
| 337 | |a ohne Hilfsmittel zu benutzen |b n |2 rdamedia | ||
| 338 | |a Band |b nc |2 rdacarrier | ||
| 520 | |a Positron emission tomography (PET) images usually suffer from a noticeable amount of statistical noise. In order to reduce this noise, a post-filtering process is usually adopted. However, the performance of this approach is limited because the denoising process is mostly performed on the basis of the Gaussian random noise. It has been reported that in a PET image reconstructed by the expectation-maximization (EM), the noise variance of each voxel depends on its mean value, unlike in the case of Gaussian noise. In addition, we observe that the variance also varies with the spatial sensitivity distribution in a PET system, which reflects both the solid angle determined by a given scanner geometry and the attenuation information of a scanned object. Thus, if a post-filtering process based on the Gaussian random noise is applied to PET images without consideration of the noise characteristics along with the spatial sensitivity distribution, the spatially variant non-Gaussian noise cannot be reduced effectively. In the proposed framework, to effectively reduce the noise in PET images reconstructed by the 3-D ordinary Poisson ordered subset EM (3-D OP-OSEM), we first denormalize an image according to the sensitivity of each voxel so that the voxel mean value can represent its statistical properties reliably. Based on our observation that each noisy denormalized voxel has a linear relationship between the mean and variance, we try to convert this non-Gaussian noise image to a Gaussian noise image. We then apply a block matching 4-D algorithm that is optimized for noise reduction of the Gaussian noise image, and reconvert and renormalize the result to obtain a final denoised image. Using simulated phantom data and clinical patient data, we demonstrate that the proposed framework can effectively suppress the noise over the whole region of a PET image while minimizing degradation of the image resolution. | ||
| 650 | 4 | |a nonGaussian noise image | |
| 650 | 4 | |a postfiltering framework | |
| 650 | 4 | |a image matching | |
| 650 | 4 | |a clinical patient data | |
| 650 | 4 | |a 3D ordinary Poisson ordered subset EM | |
| 650 | 4 | |a statistical noise | |
| 650 | 4 | |a positron emission tomography images | |
| 650 | 4 | |a positron emission tomography (PET) | |
| 650 | 4 | |a non-Gaussian noise | |
| 650 | 4 | |a noise characteristics | |
| 650 | 4 | |a PET system | |
| 650 | 4 | |a expectation-maximisation algorithm | |
| 650 | 4 | |a voxel sensitivity | |
| 650 | 4 | |a expectation-maximization | |
| 650 | 4 | |a image filtering | |
| 650 | 4 | |a sensitivity characteristics | |
| 650 | 4 | |a statistical properties | |
| 650 | 4 | |a image reconstruction | |
| 650 | 4 | |a Gaussian noise | |
| 650 | 4 | |a Gaussian random noise | |
| 650 | 4 | |a noise reduction | |
| 650 | 4 | |a postfiltering process | |
| 650 | 4 | |a Sensitivity | |
| 650 | 4 | |a noisy denormalized voxel | |
| 650 | 4 | |a Monte Carlo methods | |
| 650 | 4 | |a 3D PET image denoising | |
| 650 | 4 | |a phantom data | |
| 650 | 4 | |a Attenuation | |
| 650 | 4 | |a positron emission tomography | |
| 650 | 4 | |a medical image processing | |
| 650 | 4 | |a noise characteristics conversion | |
| 650 | 4 | |a spatially variant nonGaussian noise | |
| 650 | 4 | |a block matching 4D algorithm | |
| 650 | 4 | |a spatial sensitivity distribution | |
| 650 | 4 | |a noise variance | |
| 650 | 4 | |a Noise | |
| 650 | 4 | |a denoising process | |
| 650 | 4 | |a image denoising | |
| 650 | 4 | |a Monte Carlo method | |
| 650 | 4 | |a Random noise theory | |
| 650 | 4 | |a PET imaging | |
| 650 | 4 | |a Noise control | |
| 650 | 4 | |a Analysis | |
| 700 | 0 | |a Il Jun Ahn |4 oth | |
| 700 | 0 | |a Woo Hyun Nam |4 oth | |
| 700 | 0 | |a Jong Beom Ra |4 oth | |
| 773 | 0 | 8 | |i Enthalten in |t IEEE transactions on nuclear science |d New York, NY : IEEE, 1963 |g 62(2015), 1, Seite 137-147 |w (DE-627)129547352 |w (DE-600)218510-6 |w (DE-576)014998238 |x 0018-9499 |7 nnns |
| 773 | 1 | 8 | |g volume:62 |g year:2015 |g number:1 |g pages:137-147 |
| 856 | 4 | 1 | |u http://dx.doi.org/10.1109/TNS.2014.2360176 |3 Volltext |
| 856 | 4 | 2 | |u http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6926862 |
| 912 | |a GBV_USEFLAG_A | ||
| 912 | |a SYSFLAG_A | ||
| 912 | |a GBV_OLC | ||
| 912 | |a SSG-OLC-TEC | ||
| 912 | |a SSG-OLC-PHY | ||
| 912 | |a SSG-OLC-PHA | ||
| 912 | |a GBV_ILN_70 | ||
| 951 | |a AR | ||
| 952 | |d 62 |j 2015 |e 1 |h 137-147 | ||
| author_variant |
j h k jhk |
|---|---|
| matchkey_str |
article:00189499:2015----::nfetvpsfleigrmwrfrdeiaeeosnbsdnosade |
| hierarchy_sort_str |
2015 |
| publishDate |
2015 |
| allfields |
10.1109/TNS.2014.2360176 doi PQ20160617 (DE-627)OLC196622365X (DE-599)GBVOLC196622365X (PRQ)c1776-ac9a27912cb4c6ff7cd557cb2c8263ed56b68ec02f1e24b8d28d5d59e2b9f7880 (KEY)0054996720150000062000100137effectivepostfilteringframeworkfor3dpetimagedenois DE-627 ger DE-627 rakwb eng 620 DNB Ji Hye Kim verfasserin aut An Effective Post-Filtering Framework for 3-D PET Image Denoising Based on Noise and Sensitivity Characteristics 2015 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier Positron emission tomography (PET) images usually suffer from a noticeable amount of statistical noise. In order to reduce this noise, a post-filtering process is usually adopted. However, the performance of this approach is limited because the denoising process is mostly performed on the basis of the Gaussian random noise. It has been reported that in a PET image reconstructed by the expectation-maximization (EM), the noise variance of each voxel depends on its mean value, unlike in the case of Gaussian noise. In addition, we observe that the variance also varies with the spatial sensitivity distribution in a PET system, which reflects both the solid angle determined by a given scanner geometry and the attenuation information of a scanned object. Thus, if a post-filtering process based on the Gaussian random noise is applied to PET images without consideration of the noise characteristics along with the spatial sensitivity distribution, the spatially variant non-Gaussian noise cannot be reduced effectively. In the proposed framework, to effectively reduce the noise in PET images reconstructed by the 3-D ordinary Poisson ordered subset EM (3-D OP-OSEM), we first denormalize an image according to the sensitivity of each voxel so that the voxel mean value can represent its statistical properties reliably. Based on our observation that each noisy denormalized voxel has a linear relationship between the mean and variance, we try to convert this non-Gaussian noise image to a Gaussian noise image. We then apply a block matching 4-D algorithm that is optimized for noise reduction of the Gaussian noise image, and reconvert and renormalize the result to obtain a final denoised image. Using simulated phantom data and clinical patient data, we demonstrate that the proposed framework can effectively suppress the noise over the whole region of a PET image while minimizing degradation of the image resolution. nonGaussian noise image postfiltering framework image matching clinical patient data 3D ordinary Poisson ordered subset EM statistical noise positron emission tomography images positron emission tomography (PET) non-Gaussian noise noise characteristics PET system expectation-maximisation algorithm voxel sensitivity expectation-maximization image filtering sensitivity characteristics statistical properties image reconstruction Gaussian noise Gaussian random noise noise reduction postfiltering process Sensitivity noisy denormalized voxel Monte Carlo methods 3D PET image denoising phantom data Attenuation positron emission tomography medical image processing noise characteristics conversion spatially variant nonGaussian noise block matching 4D algorithm spatial sensitivity distribution noise variance Noise denoising process image denoising Monte Carlo method Random noise theory PET imaging Noise control Analysis Il Jun Ahn oth Woo Hyun Nam oth Jong Beom Ra oth Enthalten in IEEE transactions on nuclear science New York, NY : IEEE, 1963 62(2015), 1, Seite 137-147 (DE-627)129547352 (DE-600)218510-6 (DE-576)014998238 0018-9499 nnns volume:62 year:2015 number:1 pages:137-147 http://dx.doi.org/10.1109/TNS.2014.2360176 Volltext http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6926862 GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-TEC SSG-OLC-PHY SSG-OLC-PHA GBV_ILN_70 AR 62 2015 1 137-147 |
| spelling |
10.1109/TNS.2014.2360176 doi PQ20160617 (DE-627)OLC196622365X (DE-599)GBVOLC196622365X (PRQ)c1776-ac9a27912cb4c6ff7cd557cb2c8263ed56b68ec02f1e24b8d28d5d59e2b9f7880 (KEY)0054996720150000062000100137effectivepostfilteringframeworkfor3dpetimagedenois DE-627 ger DE-627 rakwb eng 620 DNB Ji Hye Kim verfasserin aut An Effective Post-Filtering Framework for 3-D PET Image Denoising Based on Noise and Sensitivity Characteristics 2015 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier Positron emission tomography (PET) images usually suffer from a noticeable amount of statistical noise. In order to reduce this noise, a post-filtering process is usually adopted. However, the performance of this approach is limited because the denoising process is mostly performed on the basis of the Gaussian random noise. It has been reported that in a PET image reconstructed by the expectation-maximization (EM), the noise variance of each voxel depends on its mean value, unlike in the case of Gaussian noise. In addition, we observe that the variance also varies with the spatial sensitivity distribution in a PET system, which reflects both the solid angle determined by a given scanner geometry and the attenuation information of a scanned object. Thus, if a post-filtering process based on the Gaussian random noise is applied to PET images without consideration of the noise characteristics along with the spatial sensitivity distribution, the spatially variant non-Gaussian noise cannot be reduced effectively. In the proposed framework, to effectively reduce the noise in PET images reconstructed by the 3-D ordinary Poisson ordered subset EM (3-D OP-OSEM), we first denormalize an image according to the sensitivity of each voxel so that the voxel mean value can represent its statistical properties reliably. Based on our observation that each noisy denormalized voxel has a linear relationship between the mean and variance, we try to convert this non-Gaussian noise image to a Gaussian noise image. We then apply a block matching 4-D algorithm that is optimized for noise reduction of the Gaussian noise image, and reconvert and renormalize the result to obtain a final denoised image. Using simulated phantom data and clinical patient data, we demonstrate that the proposed framework can effectively suppress the noise over the whole region of a PET image while minimizing degradation of the image resolution. nonGaussian noise image postfiltering framework image matching clinical patient data 3D ordinary Poisson ordered subset EM statistical noise positron emission tomography images positron emission tomography (PET) non-Gaussian noise noise characteristics PET system expectation-maximisation algorithm voxel sensitivity expectation-maximization image filtering sensitivity characteristics statistical properties image reconstruction Gaussian noise Gaussian random noise noise reduction postfiltering process Sensitivity noisy denormalized voxel Monte Carlo methods 3D PET image denoising phantom data Attenuation positron emission tomography medical image processing noise characteristics conversion spatially variant nonGaussian noise block matching 4D algorithm spatial sensitivity distribution noise variance Noise denoising process image denoising Monte Carlo method Random noise theory PET imaging Noise control Analysis Il Jun Ahn oth Woo Hyun Nam oth Jong Beom Ra oth Enthalten in IEEE transactions on nuclear science New York, NY : IEEE, 1963 62(2015), 1, Seite 137-147 (DE-627)129547352 (DE-600)218510-6 (DE-576)014998238 0018-9499 nnns volume:62 year:2015 number:1 pages:137-147 http://dx.doi.org/10.1109/TNS.2014.2360176 Volltext http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6926862 GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-TEC SSG-OLC-PHY SSG-OLC-PHA GBV_ILN_70 AR 62 2015 1 137-147 |
| allfields_unstemmed |
10.1109/TNS.2014.2360176 doi PQ20160617 (DE-627)OLC196622365X (DE-599)GBVOLC196622365X (PRQ)c1776-ac9a27912cb4c6ff7cd557cb2c8263ed56b68ec02f1e24b8d28d5d59e2b9f7880 (KEY)0054996720150000062000100137effectivepostfilteringframeworkfor3dpetimagedenois DE-627 ger DE-627 rakwb eng 620 DNB Ji Hye Kim verfasserin aut An Effective Post-Filtering Framework for 3-D PET Image Denoising Based on Noise and Sensitivity Characteristics 2015 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier Positron emission tomography (PET) images usually suffer from a noticeable amount of statistical noise. In order to reduce this noise, a post-filtering process is usually adopted. However, the performance of this approach is limited because the denoising process is mostly performed on the basis of the Gaussian random noise. It has been reported that in a PET image reconstructed by the expectation-maximization (EM), the noise variance of each voxel depends on its mean value, unlike in the case of Gaussian noise. In addition, we observe that the variance also varies with the spatial sensitivity distribution in a PET system, which reflects both the solid angle determined by a given scanner geometry and the attenuation information of a scanned object. Thus, if a post-filtering process based on the Gaussian random noise is applied to PET images without consideration of the noise characteristics along with the spatial sensitivity distribution, the spatially variant non-Gaussian noise cannot be reduced effectively. In the proposed framework, to effectively reduce the noise in PET images reconstructed by the 3-D ordinary Poisson ordered subset EM (3-D OP-OSEM), we first denormalize an image according to the sensitivity of each voxel so that the voxel mean value can represent its statistical properties reliably. Based on our observation that each noisy denormalized voxel has a linear relationship between the mean and variance, we try to convert this non-Gaussian noise image to a Gaussian noise image. We then apply a block matching 4-D algorithm that is optimized for noise reduction of the Gaussian noise image, and reconvert and renormalize the result to obtain a final denoised image. Using simulated phantom data and clinical patient data, we demonstrate that the proposed framework can effectively suppress the noise over the whole region of a PET image while minimizing degradation of the image resolution. nonGaussian noise image postfiltering framework image matching clinical patient data 3D ordinary Poisson ordered subset EM statistical noise positron emission tomography images positron emission tomography (PET) non-Gaussian noise noise characteristics PET system expectation-maximisation algorithm voxel sensitivity expectation-maximization image filtering sensitivity characteristics statistical properties image reconstruction Gaussian noise Gaussian random noise noise reduction postfiltering process Sensitivity noisy denormalized voxel Monte Carlo methods 3D PET image denoising phantom data Attenuation positron emission tomography medical image processing noise characteristics conversion spatially variant nonGaussian noise block matching 4D algorithm spatial sensitivity distribution noise variance Noise denoising process image denoising Monte Carlo method Random noise theory PET imaging Noise control Analysis Il Jun Ahn oth Woo Hyun Nam oth Jong Beom Ra oth Enthalten in IEEE transactions on nuclear science New York, NY : IEEE, 1963 62(2015), 1, Seite 137-147 (DE-627)129547352 (DE-600)218510-6 (DE-576)014998238 0018-9499 nnns volume:62 year:2015 number:1 pages:137-147 http://dx.doi.org/10.1109/TNS.2014.2360176 Volltext http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6926862 GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-TEC SSG-OLC-PHY SSG-OLC-PHA GBV_ILN_70 AR 62 2015 1 137-147 |
| allfieldsGer |
10.1109/TNS.2014.2360176 doi PQ20160617 (DE-627)OLC196622365X (DE-599)GBVOLC196622365X (PRQ)c1776-ac9a27912cb4c6ff7cd557cb2c8263ed56b68ec02f1e24b8d28d5d59e2b9f7880 (KEY)0054996720150000062000100137effectivepostfilteringframeworkfor3dpetimagedenois DE-627 ger DE-627 rakwb eng 620 DNB Ji Hye Kim verfasserin aut An Effective Post-Filtering Framework for 3-D PET Image Denoising Based on Noise and Sensitivity Characteristics 2015 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier Positron emission tomography (PET) images usually suffer from a noticeable amount of statistical noise. In order to reduce this noise, a post-filtering process is usually adopted. However, the performance of this approach is limited because the denoising process is mostly performed on the basis of the Gaussian random noise. It has been reported that in a PET image reconstructed by the expectation-maximization (EM), the noise variance of each voxel depends on its mean value, unlike in the case of Gaussian noise. In addition, we observe that the variance also varies with the spatial sensitivity distribution in a PET system, which reflects both the solid angle determined by a given scanner geometry and the attenuation information of a scanned object. Thus, if a post-filtering process based on the Gaussian random noise is applied to PET images without consideration of the noise characteristics along with the spatial sensitivity distribution, the spatially variant non-Gaussian noise cannot be reduced effectively. In the proposed framework, to effectively reduce the noise in PET images reconstructed by the 3-D ordinary Poisson ordered subset EM (3-D OP-OSEM), we first denormalize an image according to the sensitivity of each voxel so that the voxel mean value can represent its statistical properties reliably. Based on our observation that each noisy denormalized voxel has a linear relationship between the mean and variance, we try to convert this non-Gaussian noise image to a Gaussian noise image. We then apply a block matching 4-D algorithm that is optimized for noise reduction of the Gaussian noise image, and reconvert and renormalize the result to obtain a final denoised image. Using simulated phantom data and clinical patient data, we demonstrate that the proposed framework can effectively suppress the noise over the whole region of a PET image while minimizing degradation of the image resolution. nonGaussian noise image postfiltering framework image matching clinical patient data 3D ordinary Poisson ordered subset EM statistical noise positron emission tomography images positron emission tomography (PET) non-Gaussian noise noise characteristics PET system expectation-maximisation algorithm voxel sensitivity expectation-maximization image filtering sensitivity characteristics statistical properties image reconstruction Gaussian noise Gaussian random noise noise reduction postfiltering process Sensitivity noisy denormalized voxel Monte Carlo methods 3D PET image denoising phantom data Attenuation positron emission tomography medical image processing noise characteristics conversion spatially variant nonGaussian noise block matching 4D algorithm spatial sensitivity distribution noise variance Noise denoising process image denoising Monte Carlo method Random noise theory PET imaging Noise control Analysis Il Jun Ahn oth Woo Hyun Nam oth Jong Beom Ra oth Enthalten in IEEE transactions on nuclear science New York, NY : IEEE, 1963 62(2015), 1, Seite 137-147 (DE-627)129547352 (DE-600)218510-6 (DE-576)014998238 0018-9499 nnns volume:62 year:2015 number:1 pages:137-147 http://dx.doi.org/10.1109/TNS.2014.2360176 Volltext http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6926862 GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-TEC SSG-OLC-PHY SSG-OLC-PHA GBV_ILN_70 AR 62 2015 1 137-147 |
| allfieldsSound |
10.1109/TNS.2014.2360176 doi PQ20160617 (DE-627)OLC196622365X (DE-599)GBVOLC196622365X (PRQ)c1776-ac9a27912cb4c6ff7cd557cb2c8263ed56b68ec02f1e24b8d28d5d59e2b9f7880 (KEY)0054996720150000062000100137effectivepostfilteringframeworkfor3dpetimagedenois DE-627 ger DE-627 rakwb eng 620 DNB Ji Hye Kim verfasserin aut An Effective Post-Filtering Framework for 3-D PET Image Denoising Based on Noise and Sensitivity Characteristics 2015 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier Positron emission tomography (PET) images usually suffer from a noticeable amount of statistical noise. In order to reduce this noise, a post-filtering process is usually adopted. However, the performance of this approach is limited because the denoising process is mostly performed on the basis of the Gaussian random noise. It has been reported that in a PET image reconstructed by the expectation-maximization (EM), the noise variance of each voxel depends on its mean value, unlike in the case of Gaussian noise. In addition, we observe that the variance also varies with the spatial sensitivity distribution in a PET system, which reflects both the solid angle determined by a given scanner geometry and the attenuation information of a scanned object. Thus, if a post-filtering process based on the Gaussian random noise is applied to PET images without consideration of the noise characteristics along with the spatial sensitivity distribution, the spatially variant non-Gaussian noise cannot be reduced effectively. In the proposed framework, to effectively reduce the noise in PET images reconstructed by the 3-D ordinary Poisson ordered subset EM (3-D OP-OSEM), we first denormalize an image according to the sensitivity of each voxel so that the voxel mean value can represent its statistical properties reliably. Based on our observation that each noisy denormalized voxel has a linear relationship between the mean and variance, we try to convert this non-Gaussian noise image to a Gaussian noise image. We then apply a block matching 4-D algorithm that is optimized for noise reduction of the Gaussian noise image, and reconvert and renormalize the result to obtain a final denoised image. Using simulated phantom data and clinical patient data, we demonstrate that the proposed framework can effectively suppress the noise over the whole region of a PET image while minimizing degradation of the image resolution. nonGaussian noise image postfiltering framework image matching clinical patient data 3D ordinary Poisson ordered subset EM statistical noise positron emission tomography images positron emission tomography (PET) non-Gaussian noise noise characteristics PET system expectation-maximisation algorithm voxel sensitivity expectation-maximization image filtering sensitivity characteristics statistical properties image reconstruction Gaussian noise Gaussian random noise noise reduction postfiltering process Sensitivity noisy denormalized voxel Monte Carlo methods 3D PET image denoising phantom data Attenuation positron emission tomography medical image processing noise characteristics conversion spatially variant nonGaussian noise block matching 4D algorithm spatial sensitivity distribution noise variance Noise denoising process image denoising Monte Carlo method Random noise theory PET imaging Noise control Analysis Il Jun Ahn oth Woo Hyun Nam oth Jong Beom Ra oth Enthalten in IEEE transactions on nuclear science New York, NY : IEEE, 1963 62(2015), 1, Seite 137-147 (DE-627)129547352 (DE-600)218510-6 (DE-576)014998238 0018-9499 nnns volume:62 year:2015 number:1 pages:137-147 http://dx.doi.org/10.1109/TNS.2014.2360176 Volltext http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6926862 GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-TEC SSG-OLC-PHY SSG-OLC-PHA GBV_ILN_70 AR 62 2015 1 137-147 |
| language |
English |
| source |
Enthalten in IEEE transactions on nuclear science 62(2015), 1, Seite 137-147 volume:62 year:2015 number:1 pages:137-147 |
| sourceStr |
Enthalten in IEEE transactions on nuclear science 62(2015), 1, Seite 137-147 volume:62 year:2015 number:1 pages:137-147 |
| format_phy_str_mv |
Article |
| institution |
findex.gbv.de |
| topic_facet |
nonGaussian noise image postfiltering framework image matching clinical patient data 3D ordinary Poisson ordered subset EM statistical noise positron emission tomography images positron emission tomography (PET) non-Gaussian noise noise characteristics PET system expectation-maximisation algorithm voxel sensitivity expectation-maximization image filtering sensitivity characteristics statistical properties image reconstruction Gaussian noise Gaussian random noise noise reduction postfiltering process Sensitivity noisy denormalized voxel Monte Carlo methods 3D PET image denoising phantom data Attenuation positron emission tomography medical image processing noise characteristics conversion spatially variant nonGaussian noise block matching 4D algorithm spatial sensitivity distribution noise variance Noise denoising process image denoising Monte Carlo method Random noise theory PET imaging Noise control Analysis |
| dewey-raw |
620 |
| isfreeaccess_bool |
false |
| container_title |
IEEE transactions on nuclear science |
| authorswithroles_txt_mv |
Ji Hye Kim @@aut@@ Il Jun Ahn @@oth@@ Woo Hyun Nam @@oth@@ Jong Beom Ra @@oth@@ |
| publishDateDaySort_date |
2015-01-01T00:00:00Z |
| hierarchy_top_id |
129547352 |
| dewey-sort |
3620 |
| id |
OLC196622365X |
| language_de |
englisch |
| fullrecord |
<?xml version="1.0" encoding="UTF-8"?><collection xmlns="http://www.loc.gov/MARC21/slim"><record><leader>01000caa a2200265 4500</leader><controlfield tag="001">OLC196622365X</controlfield><controlfield tag="003">DE-627</controlfield><controlfield tag="005">20230525173913.0</controlfield><controlfield tag="007">tu</controlfield><controlfield tag="008">160206s2015 xx ||||| 00| ||eng c</controlfield><datafield tag="024" ind1="7" ind2=" "><subfield code="a">10.1109/TNS.2014.2360176</subfield><subfield code="2">doi</subfield></datafield><datafield tag="028" ind1="5" ind2="2"><subfield code="a">PQ20160617</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-627)OLC196622365X</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-599)GBVOLC196622365X</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(PRQ)c1776-ac9a27912cb4c6ff7cd557cb2c8263ed56b68ec02f1e24b8d28d5d59e2b9f7880</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(KEY)0054996720150000062000100137effectivepostfilteringframeworkfor3dpetimagedenois</subfield></datafield><datafield tag="040" ind1=" " ind2=" "><subfield code="a">DE-627</subfield><subfield code="b">ger</subfield><subfield code="c">DE-627</subfield><subfield code="e">rakwb</subfield></datafield><datafield tag="041" ind1=" " ind2=" "><subfield code="a">eng</subfield></datafield><datafield tag="082" ind1="0" ind2="4"><subfield code="a">620</subfield><subfield code="q">DNB</subfield></datafield><datafield tag="100" ind1="0" ind2=" "><subfield code="a">Ji Hye Kim</subfield><subfield code="e">verfasserin</subfield><subfield code="4">aut</subfield></datafield><datafield tag="245" ind1="1" ind2="3"><subfield code="a">An Effective Post-Filtering Framework for 3-D PET Image Denoising Based on Noise and Sensitivity Characteristics</subfield></datafield><datafield tag="264" ind1=" " ind2="1"><subfield code="c">2015</subfield></datafield><datafield tag="336" ind1=" " ind2=" "><subfield code="a">Text</subfield><subfield code="b">txt</subfield><subfield code="2">rdacontent</subfield></datafield><datafield tag="337" ind1=" " ind2=" "><subfield code="a">ohne Hilfsmittel zu benutzen</subfield><subfield code="b">n</subfield><subfield code="2">rdamedia</subfield></datafield><datafield tag="338" ind1=" " ind2=" "><subfield code="a">Band</subfield><subfield code="b">nc</subfield><subfield code="2">rdacarrier</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">Positron emission tomography (PET) images usually suffer from a noticeable amount of statistical noise. In order to reduce this noise, a post-filtering process is usually adopted. However, the performance of this approach is limited because the denoising process is mostly performed on the basis of the Gaussian random noise. It has been reported that in a PET image reconstructed by the expectation-maximization (EM), the noise variance of each voxel depends on its mean value, unlike in the case of Gaussian noise. In addition, we observe that the variance also varies with the spatial sensitivity distribution in a PET system, which reflects both the solid angle determined by a given scanner geometry and the attenuation information of a scanned object. Thus, if a post-filtering process based on the Gaussian random noise is applied to PET images without consideration of the noise characteristics along with the spatial sensitivity distribution, the spatially variant non-Gaussian noise cannot be reduced effectively. In the proposed framework, to effectively reduce the noise in PET images reconstructed by the 3-D ordinary Poisson ordered subset EM (3-D OP-OSEM), we first denormalize an image according to the sensitivity of each voxel so that the voxel mean value can represent its statistical properties reliably. Based on our observation that each noisy denormalized voxel has a linear relationship between the mean and variance, we try to convert this non-Gaussian noise image to a Gaussian noise image. We then apply a block matching 4-D algorithm that is optimized for noise reduction of the Gaussian noise image, and reconvert and renormalize the result to obtain a final denoised image. Using simulated phantom data and clinical patient data, we demonstrate that the proposed framework can effectively suppress the noise over the whole region of a PET image while minimizing degradation of the image resolution.</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">nonGaussian noise image</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">postfiltering framework</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">image matching</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">clinical patient data</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">3D ordinary Poisson ordered subset EM</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">statistical noise</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">positron emission tomography images</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">positron emission tomography (PET)</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">non-Gaussian noise</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">noise characteristics</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">PET system</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">expectation-maximisation algorithm</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">voxel sensitivity</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">expectation-maximization</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">image filtering</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">sensitivity characteristics</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">statistical properties</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">image reconstruction</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Gaussian noise</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Gaussian random noise</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">noise reduction</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">postfiltering process</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Sensitivity</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">noisy denormalized voxel</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Monte Carlo methods</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">3D PET image denoising</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">phantom data</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Attenuation</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">positron emission tomography</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">medical image processing</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">noise characteristics conversion</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">spatially variant nonGaussian noise</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">block matching 4D algorithm</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">spatial sensitivity distribution</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">noise variance</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Noise</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">denoising process</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">image denoising</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Monte Carlo method</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Random noise theory</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">PET imaging</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Noise control</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Analysis</subfield></datafield><datafield tag="700" ind1="0" ind2=" "><subfield code="a">Il Jun Ahn</subfield><subfield code="4">oth</subfield></datafield><datafield tag="700" ind1="0" ind2=" "><subfield code="a">Woo Hyun Nam</subfield><subfield code="4">oth</subfield></datafield><datafield tag="700" ind1="0" ind2=" "><subfield code="a">Jong Beom Ra</subfield><subfield code="4">oth</subfield></datafield><datafield tag="773" ind1="0" ind2="8"><subfield code="i">Enthalten in</subfield><subfield code="t">IEEE transactions on nuclear science</subfield><subfield code="d">New York, NY : IEEE, 1963</subfield><subfield code="g">62(2015), 1, Seite 137-147</subfield><subfield code="w">(DE-627)129547352</subfield><subfield code="w">(DE-600)218510-6</subfield><subfield code="w">(DE-576)014998238</subfield><subfield code="x">0018-9499</subfield><subfield code="7">nnns</subfield></datafield><datafield tag="773" ind1="1" ind2="8"><subfield code="g">volume:62</subfield><subfield code="g">year:2015</subfield><subfield code="g">number:1</subfield><subfield code="g">pages:137-147</subfield></datafield><datafield tag="856" ind1="4" ind2="1"><subfield code="u">http://dx.doi.org/10.1109/TNS.2014.2360176</subfield><subfield code="3">Volltext</subfield></datafield><datafield tag="856" ind1="4" ind2="2"><subfield code="u">http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6926862</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_USEFLAG_A</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SYSFLAG_A</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_OLC</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OLC-TEC</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OLC-PHY</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OLC-PHA</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_70</subfield></datafield><datafield tag="951" ind1=" " ind2=" "><subfield code="a">AR</subfield></datafield><datafield tag="952" ind1=" " ind2=" "><subfield code="d">62</subfield><subfield code="j">2015</subfield><subfield code="e">1</subfield><subfield code="h">137-147</subfield></datafield></record></collection>
|
| author |
Ji Hye Kim |
| spellingShingle |
Ji Hye Kim ddc 620 misc nonGaussian noise image misc postfiltering framework misc image matching misc clinical patient data misc 3D ordinary Poisson ordered subset EM misc statistical noise misc positron emission tomography images misc positron emission tomography (PET) misc non-Gaussian noise misc noise characteristics misc PET system misc expectation-maximisation algorithm misc voxel sensitivity misc expectation-maximization misc image filtering misc sensitivity characteristics misc statistical properties misc image reconstruction misc Gaussian noise misc Gaussian random noise misc noise reduction misc postfiltering process misc Sensitivity misc noisy denormalized voxel misc Monte Carlo methods misc 3D PET image denoising misc phantom data misc Attenuation misc positron emission tomography misc medical image processing misc noise characteristics conversion misc spatially variant nonGaussian noise misc block matching 4D algorithm misc spatial sensitivity distribution misc noise variance misc Noise misc denoising process misc image denoising misc Monte Carlo method misc Random noise theory misc PET imaging misc Noise control misc Analysis An Effective Post-Filtering Framework for 3-D PET Image Denoising Based on Noise and Sensitivity Characteristics |
| authorStr |
Ji Hye Kim |
| ppnlink_with_tag_str_mv |
@@773@@(DE-627)129547352 |
| format |
Article |
| dewey-ones |
620 - Engineering & allied operations |
| delete_txt_mv |
keep |
| author_role |
aut |
| collection |
OLC |
| remote_str |
false |
| illustrated |
Not Illustrated |
| issn |
0018-9499 |
| topic_title |
620 DNB An Effective Post-Filtering Framework for 3-D PET Image Denoising Based on Noise and Sensitivity Characteristics nonGaussian noise image postfiltering framework image matching clinical patient data 3D ordinary Poisson ordered subset EM statistical noise positron emission tomography images positron emission tomography (PET) non-Gaussian noise noise characteristics PET system expectation-maximisation algorithm voxel sensitivity expectation-maximization image filtering sensitivity characteristics statistical properties image reconstruction Gaussian noise Gaussian random noise noise reduction postfiltering process Sensitivity noisy denormalized voxel Monte Carlo methods 3D PET image denoising phantom data Attenuation positron emission tomography medical image processing noise characteristics conversion spatially variant nonGaussian noise block matching 4D algorithm spatial sensitivity distribution noise variance Noise denoising process image denoising Monte Carlo method Random noise theory PET imaging Noise control Analysis |
| topic |
ddc 620 misc nonGaussian noise image misc postfiltering framework misc image matching misc clinical patient data misc 3D ordinary Poisson ordered subset EM misc statistical noise misc positron emission tomography images misc positron emission tomography (PET) misc non-Gaussian noise misc noise characteristics misc PET system misc expectation-maximisation algorithm misc voxel sensitivity misc expectation-maximization misc image filtering misc sensitivity characteristics misc statistical properties misc image reconstruction misc Gaussian noise misc Gaussian random noise misc noise reduction misc postfiltering process misc Sensitivity misc noisy denormalized voxel misc Monte Carlo methods misc 3D PET image denoising misc phantom data misc Attenuation misc positron emission tomography misc medical image processing misc noise characteristics conversion misc spatially variant nonGaussian noise misc block matching 4D algorithm misc spatial sensitivity distribution misc noise variance misc Noise misc denoising process misc image denoising misc Monte Carlo method misc Random noise theory misc PET imaging misc Noise control misc Analysis |
| topic_unstemmed |
ddc 620 misc nonGaussian noise image misc postfiltering framework misc image matching misc clinical patient data misc 3D ordinary Poisson ordered subset EM misc statistical noise misc positron emission tomography images misc positron emission tomography (PET) misc non-Gaussian noise misc noise characteristics misc PET system misc expectation-maximisation algorithm misc voxel sensitivity misc expectation-maximization misc image filtering misc sensitivity characteristics misc statistical properties misc image reconstruction misc Gaussian noise misc Gaussian random noise misc noise reduction misc postfiltering process misc Sensitivity misc noisy denormalized voxel misc Monte Carlo methods misc 3D PET image denoising misc phantom data misc Attenuation misc positron emission tomography misc medical image processing misc noise characteristics conversion misc spatially variant nonGaussian noise misc block matching 4D algorithm misc spatial sensitivity distribution misc noise variance misc Noise misc denoising process misc image denoising misc Monte Carlo method misc Random noise theory misc PET imaging misc Noise control misc Analysis |
| topic_browse |
ddc 620 misc nonGaussian noise image misc postfiltering framework misc image matching misc clinical patient data misc 3D ordinary Poisson ordered subset EM misc statistical noise misc positron emission tomography images misc positron emission tomography (PET) misc non-Gaussian noise misc noise characteristics misc PET system misc expectation-maximisation algorithm misc voxel sensitivity misc expectation-maximization misc image filtering misc sensitivity characteristics misc statistical properties misc image reconstruction misc Gaussian noise misc Gaussian random noise misc noise reduction misc postfiltering process misc Sensitivity misc noisy denormalized voxel misc Monte Carlo methods misc 3D PET image denoising misc phantom data misc Attenuation misc positron emission tomography misc medical image processing misc noise characteristics conversion misc spatially variant nonGaussian noise misc block matching 4D algorithm misc spatial sensitivity distribution misc noise variance misc Noise misc denoising process misc image denoising misc Monte Carlo method misc Random noise theory misc PET imaging misc Noise control misc Analysis |
| format_facet |
Aufsätze Gedruckte Aufsätze |
| format_main_str_mv |
Text Zeitschrift/Artikel |
| carriertype_str_mv |
nc |
| author2_variant |
i j a ija w h n whn j b r jbr |
| hierarchy_parent_title |
IEEE transactions on nuclear science |
| hierarchy_parent_id |
129547352 |
| dewey-tens |
620 - Engineering |
| hierarchy_top_title |
IEEE transactions on nuclear science |
| isfreeaccess_txt |
false |
| familylinks_str_mv |
(DE-627)129547352 (DE-600)218510-6 (DE-576)014998238 |
| title |
An Effective Post-Filtering Framework for 3-D PET Image Denoising Based on Noise and Sensitivity Characteristics |
| ctrlnum |
(DE-627)OLC196622365X (DE-599)GBVOLC196622365X (PRQ)c1776-ac9a27912cb4c6ff7cd557cb2c8263ed56b68ec02f1e24b8d28d5d59e2b9f7880 (KEY)0054996720150000062000100137effectivepostfilteringframeworkfor3dpetimagedenois |
| title_full |
An Effective Post-Filtering Framework for 3-D PET Image Denoising Based on Noise and Sensitivity Characteristics |
| author_sort |
Ji Hye Kim |
| journal |
IEEE transactions on nuclear science |
| journalStr |
IEEE transactions on nuclear science |
| lang_code |
eng |
| isOA_bool |
false |
| dewey-hundreds |
600 - Technology |
| recordtype |
marc |
| publishDateSort |
2015 |
| contenttype_str_mv |
txt |
| container_start_page |
137 |
| author_browse |
Ji Hye Kim |
| container_volume |
62 |
| class |
620 DNB |
| format_se |
Aufsätze |
| author-letter |
Ji Hye Kim |
| doi_str_mv |
10.1109/TNS.2014.2360176 |
| dewey-full |
620 |
| title_sort |
effective post-filtering framework for 3-d pet image denoising based on noise and sensitivity characteristics |
| title_auth |
An Effective Post-Filtering Framework for 3-D PET Image Denoising Based on Noise and Sensitivity Characteristics |
| abstract |
Positron emission tomography (PET) images usually suffer from a noticeable amount of statistical noise. In order to reduce this noise, a post-filtering process is usually adopted. However, the performance of this approach is limited because the denoising process is mostly performed on the basis of the Gaussian random noise. It has been reported that in a PET image reconstructed by the expectation-maximization (EM), the noise variance of each voxel depends on its mean value, unlike in the case of Gaussian noise. In addition, we observe that the variance also varies with the spatial sensitivity distribution in a PET system, which reflects both the solid angle determined by a given scanner geometry and the attenuation information of a scanned object. Thus, if a post-filtering process based on the Gaussian random noise is applied to PET images without consideration of the noise characteristics along with the spatial sensitivity distribution, the spatially variant non-Gaussian noise cannot be reduced effectively. In the proposed framework, to effectively reduce the noise in PET images reconstructed by the 3-D ordinary Poisson ordered subset EM (3-D OP-OSEM), we first denormalize an image according to the sensitivity of each voxel so that the voxel mean value can represent its statistical properties reliably. Based on our observation that each noisy denormalized voxel has a linear relationship between the mean and variance, we try to convert this non-Gaussian noise image to a Gaussian noise image. We then apply a block matching 4-D algorithm that is optimized for noise reduction of the Gaussian noise image, and reconvert and renormalize the result to obtain a final denoised image. Using simulated phantom data and clinical patient data, we demonstrate that the proposed framework can effectively suppress the noise over the whole region of a PET image while minimizing degradation of the image resolution. |
| abstractGer |
Positron emission tomography (PET) images usually suffer from a noticeable amount of statistical noise. In order to reduce this noise, a post-filtering process is usually adopted. However, the performance of this approach is limited because the denoising process is mostly performed on the basis of the Gaussian random noise. It has been reported that in a PET image reconstructed by the expectation-maximization (EM), the noise variance of each voxel depends on its mean value, unlike in the case of Gaussian noise. In addition, we observe that the variance also varies with the spatial sensitivity distribution in a PET system, which reflects both the solid angle determined by a given scanner geometry and the attenuation information of a scanned object. Thus, if a post-filtering process based on the Gaussian random noise is applied to PET images without consideration of the noise characteristics along with the spatial sensitivity distribution, the spatially variant non-Gaussian noise cannot be reduced effectively. In the proposed framework, to effectively reduce the noise in PET images reconstructed by the 3-D ordinary Poisson ordered subset EM (3-D OP-OSEM), we first denormalize an image according to the sensitivity of each voxel so that the voxel mean value can represent its statistical properties reliably. Based on our observation that each noisy denormalized voxel has a linear relationship between the mean and variance, we try to convert this non-Gaussian noise image to a Gaussian noise image. We then apply a block matching 4-D algorithm that is optimized for noise reduction of the Gaussian noise image, and reconvert and renormalize the result to obtain a final denoised image. Using simulated phantom data and clinical patient data, we demonstrate that the proposed framework can effectively suppress the noise over the whole region of a PET image while minimizing degradation of the image resolution. |
| abstract_unstemmed |
Positron emission tomography (PET) images usually suffer from a noticeable amount of statistical noise. In order to reduce this noise, a post-filtering process is usually adopted. However, the performance of this approach is limited because the denoising process is mostly performed on the basis of the Gaussian random noise. It has been reported that in a PET image reconstructed by the expectation-maximization (EM), the noise variance of each voxel depends on its mean value, unlike in the case of Gaussian noise. In addition, we observe that the variance also varies with the spatial sensitivity distribution in a PET system, which reflects both the solid angle determined by a given scanner geometry and the attenuation information of a scanned object. Thus, if a post-filtering process based on the Gaussian random noise is applied to PET images without consideration of the noise characteristics along with the spatial sensitivity distribution, the spatially variant non-Gaussian noise cannot be reduced effectively. In the proposed framework, to effectively reduce the noise in PET images reconstructed by the 3-D ordinary Poisson ordered subset EM (3-D OP-OSEM), we first denormalize an image according to the sensitivity of each voxel so that the voxel mean value can represent its statistical properties reliably. Based on our observation that each noisy denormalized voxel has a linear relationship between the mean and variance, we try to convert this non-Gaussian noise image to a Gaussian noise image. We then apply a block matching 4-D algorithm that is optimized for noise reduction of the Gaussian noise image, and reconvert and renormalize the result to obtain a final denoised image. Using simulated phantom data and clinical patient data, we demonstrate that the proposed framework can effectively suppress the noise over the whole region of a PET image while minimizing degradation of the image resolution. |
| collection_details |
GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-TEC SSG-OLC-PHY SSG-OLC-PHA GBV_ILN_70 |
| container_issue |
1 |
| title_short |
An Effective Post-Filtering Framework for 3-D PET Image Denoising Based on Noise and Sensitivity Characteristics |
| url |
http://dx.doi.org/10.1109/TNS.2014.2360176 http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6926862 |
| remote_bool |
false |
| author2 |
Il Jun Ahn Woo Hyun Nam Jong Beom Ra |
| author2Str |
Il Jun Ahn Woo Hyun Nam Jong Beom Ra |
| ppnlink |
129547352 |
| mediatype_str_mv |
n |
| isOA_txt |
false |
| hochschulschrift_bool |
false |
| author2_role |
oth oth oth |
| doi_str |
10.1109/TNS.2014.2360176 |
| up_date |
2024-07-03T20:49:32.188Z |
| _version_ |
1803592424164950016 |
| fullrecord_marcxml |
<?xml version="1.0" encoding="UTF-8"?><collection xmlns="http://www.loc.gov/MARC21/slim"><record><leader>01000caa a2200265 4500</leader><controlfield tag="001">OLC196622365X</controlfield><controlfield tag="003">DE-627</controlfield><controlfield tag="005">20230525173913.0</controlfield><controlfield tag="007">tu</controlfield><controlfield tag="008">160206s2015 xx ||||| 00| ||eng c</controlfield><datafield tag="024" ind1="7" ind2=" "><subfield code="a">10.1109/TNS.2014.2360176</subfield><subfield code="2">doi</subfield></datafield><datafield tag="028" ind1="5" ind2="2"><subfield code="a">PQ20160617</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-627)OLC196622365X</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-599)GBVOLC196622365X</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(PRQ)c1776-ac9a27912cb4c6ff7cd557cb2c8263ed56b68ec02f1e24b8d28d5d59e2b9f7880</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(KEY)0054996720150000062000100137effectivepostfilteringframeworkfor3dpetimagedenois</subfield></datafield><datafield tag="040" ind1=" " ind2=" "><subfield code="a">DE-627</subfield><subfield code="b">ger</subfield><subfield code="c">DE-627</subfield><subfield code="e">rakwb</subfield></datafield><datafield tag="041" ind1=" " ind2=" "><subfield code="a">eng</subfield></datafield><datafield tag="082" ind1="0" ind2="4"><subfield code="a">620</subfield><subfield code="q">DNB</subfield></datafield><datafield tag="100" ind1="0" ind2=" "><subfield code="a">Ji Hye Kim</subfield><subfield code="e">verfasserin</subfield><subfield code="4">aut</subfield></datafield><datafield tag="245" ind1="1" ind2="3"><subfield code="a">An Effective Post-Filtering Framework for 3-D PET Image Denoising Based on Noise and Sensitivity Characteristics</subfield></datafield><datafield tag="264" ind1=" " ind2="1"><subfield code="c">2015</subfield></datafield><datafield tag="336" ind1=" " ind2=" "><subfield code="a">Text</subfield><subfield code="b">txt</subfield><subfield code="2">rdacontent</subfield></datafield><datafield tag="337" ind1=" " ind2=" "><subfield code="a">ohne Hilfsmittel zu benutzen</subfield><subfield code="b">n</subfield><subfield code="2">rdamedia</subfield></datafield><datafield tag="338" ind1=" " ind2=" "><subfield code="a">Band</subfield><subfield code="b">nc</subfield><subfield code="2">rdacarrier</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">Positron emission tomography (PET) images usually suffer from a noticeable amount of statistical noise. In order to reduce this noise, a post-filtering process is usually adopted. However, the performance of this approach is limited because the denoising process is mostly performed on the basis of the Gaussian random noise. It has been reported that in a PET image reconstructed by the expectation-maximization (EM), the noise variance of each voxel depends on its mean value, unlike in the case of Gaussian noise. In addition, we observe that the variance also varies with the spatial sensitivity distribution in a PET system, which reflects both the solid angle determined by a given scanner geometry and the attenuation information of a scanned object. Thus, if a post-filtering process based on the Gaussian random noise is applied to PET images without consideration of the noise characteristics along with the spatial sensitivity distribution, the spatially variant non-Gaussian noise cannot be reduced effectively. In the proposed framework, to effectively reduce the noise in PET images reconstructed by the 3-D ordinary Poisson ordered subset EM (3-D OP-OSEM), we first denormalize an image according to the sensitivity of each voxel so that the voxel mean value can represent its statistical properties reliably. Based on our observation that each noisy denormalized voxel has a linear relationship between the mean and variance, we try to convert this non-Gaussian noise image to a Gaussian noise image. We then apply a block matching 4-D algorithm that is optimized for noise reduction of the Gaussian noise image, and reconvert and renormalize the result to obtain a final denoised image. Using simulated phantom data and clinical patient data, we demonstrate that the proposed framework can effectively suppress the noise over the whole region of a PET image while minimizing degradation of the image resolution.</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">nonGaussian noise image</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">postfiltering framework</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">image matching</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">clinical patient data</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">3D ordinary Poisson ordered subset EM</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">statistical noise</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">positron emission tomography images</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">positron emission tomography (PET)</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">non-Gaussian noise</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">noise characteristics</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">PET system</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">expectation-maximisation algorithm</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">voxel sensitivity</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">expectation-maximization</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">image filtering</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">sensitivity characteristics</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">statistical properties</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">image reconstruction</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Gaussian noise</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Gaussian random noise</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">noise reduction</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">postfiltering process</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Sensitivity</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">noisy denormalized voxel</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Monte Carlo methods</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">3D PET image denoising</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">phantom data</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Attenuation</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">positron emission tomography</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">medical image processing</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">noise characteristics conversion</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">spatially variant nonGaussian noise</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">block matching 4D algorithm</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">spatial sensitivity distribution</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">noise variance</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Noise</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">denoising process</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">image denoising</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Monte Carlo method</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Random noise theory</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">PET imaging</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Noise control</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Analysis</subfield></datafield><datafield tag="700" ind1="0" ind2=" "><subfield code="a">Il Jun Ahn</subfield><subfield code="4">oth</subfield></datafield><datafield tag="700" ind1="0" ind2=" "><subfield code="a">Woo Hyun Nam</subfield><subfield code="4">oth</subfield></datafield><datafield tag="700" ind1="0" ind2=" "><subfield code="a">Jong Beom Ra</subfield><subfield code="4">oth</subfield></datafield><datafield tag="773" ind1="0" ind2="8"><subfield code="i">Enthalten in</subfield><subfield code="t">IEEE transactions on nuclear science</subfield><subfield code="d">New York, NY : IEEE, 1963</subfield><subfield code="g">62(2015), 1, Seite 137-147</subfield><subfield code="w">(DE-627)129547352</subfield><subfield code="w">(DE-600)218510-6</subfield><subfield code="w">(DE-576)014998238</subfield><subfield code="x">0018-9499</subfield><subfield code="7">nnns</subfield></datafield><datafield tag="773" ind1="1" ind2="8"><subfield code="g">volume:62</subfield><subfield code="g">year:2015</subfield><subfield code="g">number:1</subfield><subfield code="g">pages:137-147</subfield></datafield><datafield tag="856" ind1="4" ind2="1"><subfield code="u">http://dx.doi.org/10.1109/TNS.2014.2360176</subfield><subfield code="3">Volltext</subfield></datafield><datafield tag="856" ind1="4" ind2="2"><subfield code="u">http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6926862</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_USEFLAG_A</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SYSFLAG_A</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_OLC</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OLC-TEC</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OLC-PHY</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OLC-PHA</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_70</subfield></datafield><datafield tag="951" ind1=" " ind2=" "><subfield code="a">AR</subfield></datafield><datafield tag="952" ind1=" " ind2=" "><subfield code="d">62</subfield><subfield code="j">2015</subfield><subfield code="e">1</subfield><subfield code="h">137-147</subfield></datafield></record></collection>
|
| score |
7.3986073 |