Practical considerations for noise power spectra estimation for clinical CT scanners

Dolly, S; Chen, Hsin Chen; Anastasio, Mark A.; Mutic, Sasa; Li, Hua

doi:10.1120/jacmp.v17i3.5841

Cited by 44 publications

(51 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The 2D NPS, NPS ( u,v ) of an image ROI, I ( x,y ), was calculated as

N P S false(u, v false) = \frac{Δ x Δ y}{N_{x} N_{y}} {| F T false[I (x, y) - \overset{I}{true} false] |}^{2}

using previously published methods, where u , v represent spatial frequency (mm −1 ) in the x and y direction, respectively. N x and N y are the number of pixels in the x and y direction and Δ x and Δ y are the pixel dimensions (mm).…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Application of fractal dimension for quantifying noise texture in computed tomography images

Khobragade

Fan²,

Rupcich³

et al. 2018

Medical Physics

View full text Add to dashboard Cite

Fractal dimension correlated with the NPS-peak frequency and was independent of noise magnitude, suggesting that the scalar metric of fractal dimension can be used to quantify the change in noise texture across reconstruction approaches. Results demonstrated that fractal dimension can be estimated from four, 64 × 64-pixel ROIs or one 128 × 128 ROI within a head CT image, which may make it amenable for quantifying noise texture within clinical images.

show abstract

“…The 2D NPS, NPS ( u,v ) of an image ROI, I ( x,y ), was calculated as

N P S false(u, v false) = \frac{Δ x Δ y}{N_{x} N_{y}} {| F T false[I (x, y) - \overset{I}{true} false] |}^{2}

Section: Methodsmentioning

confidence: 99%

“…Variability in the local NPS has been observed depending on the estimation parameters such as ROI size and number of ROIs . For NPS , increasing the ROI size increases the spatial frequency resolution of the NPS ; however, fewer ROIs may be available for NPS calculation, which can lead to errors.…”

Section: Methodsmentioning

confidence: 99%

Application of fractal dimension for quantifying noise texture in computed tomography images

Khobragade

Fan²,

Rupcich³

et al. 2018

Medical Physics

View full text Add to dashboard Cite

show abstract

“…The ROI mean μ and ROI standard deviation (SD) σ of the pixels in each ROI were calculated based on the following equations:

μ = \frac{1}{N} {false\sum}_{i = 1}^{N} g_{i}

σ = \sqrt{\frac{1}{N - 1} {false\sum}_{i = 1}^{N} {1.2em1.2emtrue| g_{i} - μ 1.2em1.2emtrue|}^{2}}

where N is the total number of pixels in an ROI and g i is an individual pixel element . The two‐dimensional (2D) noise power spectrum (NPS) was calculated for each ROI g , using the following equation:

N P S (f_{i}) = \frac{p_{x} p_{y}}{N_{x} N_{y}} | DFT \{g - \bar{boldg}\} {1.2em1.2emtrue|}^{2}

where p x and p y are the pixel sizes in millimeter and N x and N y are the number of pixels in each dimension.…”

Section: Methodsmentioning

confidence: 99%

“…The two‐dimensional (2D) noise power spectrum (NPS) was calculated for each ROI g , using the following equation:

N P S (f_{i}) = \frac{p_{x} p_{y}}{N_{x} N_{y}} | DFT \{g - \bar{boldg}\} {1.2em1.2emtrue|}^{2}

where p x and p y are the pixel sizes in millimeter and N x and N y are the number of pixels in each dimension. Here,

\bar{g}

is the structured, nonstochastic background which was estimated using a 2D first‐order polynomial fit function . The aforementioned method can be extended to a second‐order polynomial fitting, with the function being defined as:

{\bar{g}}_{m, n}^{'} = C_{1} + C_{2} m + C_{3} n + C_{4} m^{2} + C_{5} n^{2} + C_{6} m n

…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Technical Note: A feasibility study of using the flat panel detector on linac for the kV x‐ray generator test

et al. 2018

Self Cite

View full text Add to dashboard Cite

show abstract

Quantitation of clinical feedback on image quality differences between two CT scanner models

Bache

Stauduhar

Liu

et al. 2017

J Applied Clin Med Phys

View full text Add to dashboard Cite

The aim of this work was to quantitate differences in image quality between two GE CT scanner models — the LightSpeed VCT (“VCT”) and Discovery HD750 (“HD”) — based upon feedback from radiologists at our institution. First, 3 yrs of daily QC images of the manufacturer‐provided QC phantom from 10 scanners — five of each model — were analyzed for both noise magnitude, measured as CT‐number standard deviation, and noise power spectrum within the uniform water section. The same phantom was then scanned on four of each model and analyzed for low contrast detectability (LCD) using a built‐in LCD tool at the scanner console. An anthropomorphic phantom was scanned using the same eight scanners. A slice within the abdomen section was chosen and three ROIs were placed in regions representing liver, stomach, and spleen. Both standard deviation of CT‐number and LCD value was calculated for each image. Noise magnitude was 8.5% higher in HD scanners compared to VCT scanners. An associated increase in the magnitude of the noise power spectra were also found, but both peak and mean NPS frequency were not different between the two models. VCT scanners outperformed HD scanners with respect to LCD by an average of 13.1% across all scanners and phantoms. Our results agree with radiologist feedback, and necessitate a closer look at our body CT protocols among different scanner models at our institution.

show abstract

Practical considerations for noise power spectra estimation for clinical CT scanners

Cited by 44 publications

References 30 publications

Application of fractal dimension for quantifying noise texture in computed tomography images

Application of fractal dimension for quantifying noise texture in computed tomography images

Technical Note: A feasibility study of using the flat panel detector on linac for the kV x‐ray generator test

Quantitation of clinical feedback on image quality differences between two CT scanner models

Contact Info

Product

Resources

About