1887
banner image
No data available.
Please log in to see this content.
You have no subscription access to this content.
No metrics data to plot.
The attempt to load metrics for this article has failed.
The attempt to plot a graph for these metrics has failed.
A full-spectral Bayesian reconstruction approach based on the material decomposition model applied in dual-energy computed tomography
Rent:
Rent this article for
USD
10.1118/1.4820478
/content/aapm/journal/medphys/40/11/10.1118/1.4820478
http://aip.metastore.ingenta.com/content/aapm/journal/medphys/40/11/10.1118/1.4820478

Figures

Image of FIG. 1.
FIG. 1.

Discrete DECT forward model with energy sampling bins.

Image of FIG. 2.
FIG. 2.

Schematic representation of Phantom 1.

Image of FIG. 3.
FIG. 3.

MSE versus PSNR of different reconstruction approaches.

Image of FIG. 4.
FIG. 4.

Reconstructed water and bone fractions from simulated projections at PSNR = 45 dB. The first row shows the PE fractions while the second shows the HA fractions. (a) True fractions; (b) DE-PWLS (δ = 0.02, σ = 0.05), MSE = 4.5 × 10−2; (c) DE-FBP with BHA correction (Hanning filter cut frequency = 0.25 times of Nyquist frequency), MSE = 6.1 × 10−3; (d) DE-Bayesian (δ = 0.02, σ = 0.05), MSE = 1.2 × 10−3.

Image of FIG. 5.
FIG. 5.

Profiles of reconstructed water and bone fractions with simulated projections at PSNR = 45 dB. Profile positions are marked by arrows from left to right on fraction images shown on the top.

Image of FIG. 6.
FIG. 6.

Estimated LACs (in dashed lines) compared with true LACs (in solid lines) for materials contained in Phantom 1 (LAC curves of salt water and HA are overlapped completely with their true LACs).

Image of FIG. 7.
FIG. 7.

DE-Bayesian results with mismatched spectrum (profile positions are marked by arrows from left to right on fraction images). Projections are simulated at 70 kV and 150 kV with PSNR = 45 dB. A 1 mm Al filter is added at 70 kV for the spectrum used in the reconstruction.

Image of FIG. 8.
FIG. 8.

Photo of Phantom 2, a piece of spinal column in resin.

Image of FIG. 9.
FIG. 9.

Experimental results of Phantom 1. The first row shows the water fractions while the second shows the bone fractions. (a) DE-PWLS (δ = 0.02, σ = 0.05); (b) DE-FBP with BHA correction (Hanning filter cut frequency = 0.25 times of Nyquist frequency); (c) DE-Bayesian (δ = 0.02, σ = 0.05).

Image of FIG. 10.
FIG. 10.

Profiles of reconstructed water and bone fractions from experimental data at 70 kV and 150 kV. Profile positions are marked by arrows on fraction images shown in the corner.

Image of FIG. 11.
FIG. 11.

Experimental results of Phantom 2. The first row shows the water fractions while the second shows the water fractions. (a) DE-PWLS (δ = 0.02, σ = 0.05); (b) DE-FBP with BHA correction (Hanning filter cut frequency = 0.25 times of Nyquist frequency); (c) DE-Bayesian (δ = 0.02, σ = 0.05).

Image of FIG. 12.
FIG. 12.

Decomposition fraction errors obtained from experimental data at PSNR ≈ 45 dB where the fraction images are obtained by applying the DE-FBP approach with BHA correction on the experimental data at PSNR ≈ 65 dB. (a) DE-PWLS; (b) DE-FBP with BHA correction; (c) DE-Bayesian.

Tables

Generic image for table
TABLE I.

CG minimization algorithm schema with suboptimal descent steps.

Generic image for table
TABLE II.

Water and bone decomposition fractions of materials contained in Phantom 1.

Generic image for table
TABLE III.

Estimated water and bone fractions from simulated results at PSNR = 45 dB.

Generic image for table
TABLE IV.

Estimated water and bone fractions with mismatched spectra (true decomposition fractions are shown in the head line).

Generic image for table
TABLE V.

Estimated water and bone fractions from experimental data for materials contained in Phantom 1.

Loading

Article metrics loading...

/content/aapm/journal/medphys/40/11/10.1118/1.4820478
2013-10-22
2014-04-19
Loading

Full text loading...

This is a required field
Please enter a valid email address
752b84549af89a08dbdd7fdb8b9568b5 journal.articlezxybnytfddd
Scitation: A full-spectral Bayesian reconstruction approach based on the material decomposition model applied in dual-energy computed tomography
http://aip.metastore.ingenta.com/content/aapm/journal/medphys/40/11/10.1118/1.4820478
10.1118/1.4820478
SEARCH_EXPAND_ITEM