Noise treatment of low-dose computed tomography projections and images

1. A computer-implemented method for treating noise in low-dose computed tomography projections and reconstructed images comprising the steps of:

acquiring raw data at a low mA value;

applying a domain specific noise filter in a sinogram domain of the raw data; and

applying an edge preserving smoothing filter in an image domain of the raw data after filtering in the sinogram domain, wherein the adaptive edge preserving smoothing filter is based on one of a curve and a fitted curve.

2. The computer-implemented method of claim 1 , wherein the low mA value is less than about 100 mA.

3. The computer-implemented method of claim 1 , wherein the raw data is patient specific raw data.

4. The computer-implemented method of claim 1 , wherein the raw data is based on anthropomorphic phantoms.

5. The computer-implemented method of claim 1 , further comprising the steps of:

determining the domain specific noise filter with the raw data, such that the domain specific noise filter has prior knowledge of a noise property of the sinogram domain prior to being applied; and

determining the edge preserving smoothing filter with the raw data.

6. The computer-implemented method of claim 1 , further comprising the steps of:

generating a curve for variances and means given the raw data;

fitting the curve by a functional form;

determining, for a fitted curve, a transformed space having substantially constant variance for all means;

determining, for a substantially constant variance, a transform space having estimated signal-to-noise ratios; and

filtering the raw data in the transformed space.

7. The computer-implemented method of claim 6 , wherein the step of filtering further comprises the step of applying one of a Wiener filter and a Kalman filter.

8. The computer-implemented method of claim 6 , wherein the fitting is one of linear, quadratic, non-linear, and square root.

9. The computer-implemented method of claim 6 , wherein the step of determining, for a fitted curve, a transformed space is by a logarithmic transform for a linear fitting, an inverse tangent transform for a quadratic fitting, a segmented logarithmic transform for a non-linear fitting, and an Anscombe transform for a square-root fitting.

10. The computer-implemented method of claim 6 , wherein the step of determining, for the substantially constant variance, a transformed space is a Karhunen-Loeve transform, whose eigenvalues and the substantially constant variance are the estimate of the signal-to-noise ratios.

11. The computer-implemented method of claim 1 , further comprising the steps of:

generating a curve for variances and means given the raw data;

fitting the curve by a functional form;

determining a signal-to-noise ratio in the transformed space;

selecting plurality of neighboring sinograms along an axial direction of the computed tomography projections, wherein the raw data is three-dimensional projection data; and

applying a Karhunen-Loeve domain penalized weighted least square smoothing filter.

12. The computer-implemented method of claim 11 , wherein determining the signal-to-noise ratio in the transformed space is by Karhunen-Loeve transform.

13. The computer-implemented method of claim 1 , further comprising the step of acquiring a point source response function for resolution enhancement by incorporating the point source response into the domain specific filters of Wiener, Kalman, and penalized weighted least square smoothing filters.

14. The computer-implemented method of claim 1 , wherein the edge preserving smoothing filter is a non-linear Gaussian filter chain, wherein the edge preserving smoothing filter is one of adaptive and non-adaptive.

15. The computer-implemented method of claim 1 , wherein the domain specific noise filter and the edge preserving smoothing filter is applied to emission computed tomography projections, where the raw data is noisy due to low dose radiopharmaceutical administration.

16. The computer-implemented method of claim 15 , comprising the steps of:

generating a curve for variances and means given the raw data;

fitting the curve by a functional form;

determining, for a fitted curve, a transformed space having substantially constant variance for all means according to an Anscombe transform; and

filtering the raw data in the transformed space, wherein a Kalman filter and a Wiener filter is applied to the raw data.

17. A computer readable medium embodying instructions executable by a processor to perform a method comprising:

generating a curve for variance and means given a set of raw data;

fitting the curve by a functional form;

determining, for a fitted curve, a transformed space having substantially constant variance for all means and having readily available signal-to-noise ratios;

applying a domain specific noise filter in a sinogram domain of the set of raw data; and

applying an edge preserving smoothing filter in an image domain of the set of raw data after filtering in the sinogram domain.

18. The method of claim 17 , wherein the computed tomography projections are emission computed tomography projections.

19. An image data filter system comprising:

a variance curve of image data:

a means curve of the image data, wherein a relationship between the variance curve and the means curve determines a fit of the image data to a minimum least-square criterion;

a domain specific noise filter receiving the image data, the domain specific noise filter for processing a sinogram domain of the image data; and

an edge preserving smoothing filter receiving the image data having a filtered sinogram domain, the edge preserving smoothing filer for processing an image domain of the image data.

20. The system of claim 19 , wherein the fit is one of a logarithmic transform for a linear fitting, an inverse tangent transform for a quadratic fitting, a segmented logarithmic transform for a non-linear fitting, and an Anscombe transform for a square-root fitting.

21. The system of claim 19 , wherein the domain specific filter is one of Wiener, Kalman, and penalized weighted least square smoothing filters.

22. The system of claim 19 , wherein the domain specific filter is adaptive to a signal-to-noise ratio of the image data.

23. The system of claim 19 , wherein the edge preserving filter is adaptive to noise properties of the image data and spatial information spatial information of the image data.

24. A computer-implemented method for treating noise in low-dose computed tomography projections and reconstructed images comprising the steps of:

acquiring raw data at a low mA value, wherein the low mA value is less than about 100 mA;

applying a domain specific noise filter in a sinogram domain of the raw data; and

applying an edge preserving smoothing filter in an image domain of the raw data after filtering in the sinogram domain.