Methods and systems for image reconstruction using low noise kernel

• US 20070196006A1
• Filed: 02/21/2006
• Published: 08/23/2007
• Est. Priority Date: 02/21/2006
• Status: Active Grant

• Current Assignee: General Electric Company
• Original Assignee: General Electric Company
• Inventors: Hsieh, Jiang
• Granted Patent: US 7,653,230 B2
• Time in Patent Office: Days
• US Class Current: 382/131
• CPC Class Codes: G06T 11/006 Inverse problem, transformation from projection-space into object-space, e.g. transform methods, back-projection, algebraic methods
• CPC Class Codes: G06T 2211/421 Filtered back projection [FBP]

Abstract

Methods and systems for a system for visualizing relatively small structures within an object are provided. The system includes an image acquisition sub-system for acquiring a dataset for a volume of interest and processor for generating image data from the acquired data wherein the processor is programmed to execute a high resolution filter kernel algorithm that includes a weighting factor applied to a ramp filter that scales relatively high frequency regions of the image dataset by a factor greater than one. The high resolution filter kernel algorithm also includes a windowing function applied to the weighted ramp filter that facilitates reducing aliasing artifacts in reconstructed images generated from the image dataset.

Claims

1. A system for visualizing relatively small structures within an object, said system comprising:

an image acquisition sub-system for acquiring a dataset for a volume of interest; and

a processor for generating image data from the acquired data, said processor programmed to execute a high resolution filter kernel algorithm comprising a weighting factor applied to a ramp filter that scales relatively high frequency regions of the image dataset by a factor greater than one, and a windowing function applied to the weighted ramp filter that facilitates reducing aliasing artifacts in reconstructed images generated from the image dataset.

2. A system in accordance with claim 1 wherein said weighting factor is a polynomial weighting factor.

3. A system in accordance with claim 1 wherein said weighting factor is configured to enhance relatively high frequency contents of the image dataset during tomographic reconstruction.

4. A system in accordance with claim 1 wherein said windowing function is configured to reduce a magnitude of aliasing frequencies to substantially zero.

5. A system in accordance with claim 1 wherein said windowing function is configured to reduce a magnitude of frequencies outside a predetermined frequency region to substantially zero.

6. A system in accordance with claim 1 wherein said windowing function is configured to reduce a magnitude of frequencies inside a selectable range of frequency to a relatively small magnitude.

7. A system in accordance with claim 6 wherein said windowing function is configured to reduce a magnitude of frequencies inside a selectable range of frequency to substantially zero.

8. A system in accordance with claim 7 wherein said selectable range of frequency is at least one of automatically selectable and selectable by a user.

9. A system in accordance with claim 7 wherein said selectable range of frequency is selectable based on a portion of an anatomy of the object.

10. A system in accordance with claim 1 wherein said windowing function is configured to maintain a shape of the weighted ramp filter inside a predetermined frequency region.

11. A system in accordance with claim 1 further comprising an image acquisition portion for acquiring image data and a controller configured to control the image acquisition portion,

12. A system in accordance with claim 1 wherein the image acquisition portion comprises at least one of an X-ray modality, a CT modality, an ultrasound modality, a magnetic resonance imaging (MRI) modality, a positron emission tomography (PET) modality, a nuclear medicine modality, an optical modality, and a combination of modalities

13. An imaging system comprising an image acquisition portion for acquiring data, a controller configured to control the image acquisition portion, and a processor configured to receive a dataset for an object that includes relatively small structures, said processor further programmed to:

process the dataset using a high resolution filter kernel algorithm comprising a weighting factor applied to a ramp filter that scales relatively high frequency regions of the image dataset by a factor greater than one, and a windowing function applied to the weighted ramp filter that facilitates reducing aliasing artifacts in reconstructed images generated from the image dataset.

14. An imaging system in accordance with claim 13 wherein said image acquisition portion comprises at least one of an X-ray source, an ultrasound source, a magnetic source, a positron emission source, a nuclear source, an optical source, and a combination of said sources;

15. An imaging system in accordance with claim 13 wherein said weighting factor is a polynomial weighting factor.

16. An imaging system in accordance with claim 13 wherein said weighting factor is configured to enhance relatively high frequency contents of the image dataset during tomographic reconstruction.

17. An imaging system in accordance with claim 13 wherein said windowing function is configured to reduce a magnitude of aliasing frequencies to substantially zero.

18. An imaging system in accordance with claim 13 wherein said windowing function is configured to reduce a magnitude of frequencies outside a Nyquist region to substantially zero.

19. An imaging system in accordance with claim 13 wherein said windowing function is configured to maintain a shape of the weighted ramp filter inside a Nyquist region.

20. A method of visualizing relatively small structures within an object comprising:

receiving a dataset for a volume of interest; and

applying a high resolution filter kernel including a weighting factor and a windowing function to the dataset wherein the weighting factor is applied to a ramp filter that scales relatively high frequency regions of the dataset by a factor greater than one and wherein the windowing function is applied to the weighted ramp filter such that aliasing artifacts in reconstructed images generated from the image dataset are facilitated being reduced.

21. A method in accordance with claim 20 wherein applying a high resolution filter kernel including a weighting factor and a windowing function to the image dataset comprises applying a high resolution filter kernel including a polynomial weighting factor.

22. A method in accordance with claim 20 wherein applying a high resolution filter kernel including a weighting factor and a windowing function to the image dataset comprises applying a high resolution filter kernel including a weighting factor that is configured to enhance relatively high frequency contents of the image dataset during tomographic reconstruction.

23. A method in accordance with claim 20 wherein applying a high resolution filter kernel including a weighting factor and a windowing function to the image dataset comprises applying a windowing function that is configured to reduce a magnitude of frequencies outside a Nyquist region to substantially zero.

24. A method in accordance with claim 20 wherein applying a high resolution filter kernel including a weighting factor and a windowing function to the image dataset comprises applying a windowing function that is configured to maintain a shape of the weighted ramp filter inside a Nyquist region.

25. A method for reconstructing an image from a dataset for a volume of interest, the method comprising:

defining a reconstruction kernel by applying a window function to a filter kernel, the reconstruction kernel configured to substantially zero out frequencies outside of a Nyquist region associated with the filter kernel while maintaining the shape of filter kernel inside the Nyquist region; and

applying the reconstruction kernel to the dataset.

26. The method of claim 25, wherein:

the filter kernel comprises an edge kernel configured to scale relatively high frequency regions of the dataset by a factor greater than one, thereby enhancing the relatively high frequency content of a projection of the dataset; and

the reconstruction kernel configured to preserve the frequencies inside the Nyquist region while suppressing aliasing artifacts associated with the zeroed out aliased high frequencies.

27. The method of claim 26, wherein:

the filter kernel comprises a weighting function applied to a ramp filter.

28. The method of claim 27 wherein the volume of interest comprises a bony structure of a patient, and further wherein:

in response to the applying the reconstruction kernel to the dataset, the sharpness of the bony structure is preserved.

29. The method of claim 28 wherein the bony structure comprises inner ear bony structure, the absence of the applying a window function enables reconstructed imaging on a display window of a first size, the presence of the applying a window function enables reconstructed imaging on a display window of a second size greater than the first size, and further wherein:

in response to the applying the reconstruction kernel to the dataset, the sharpness of the inner ear bony structure is preserved in the display window of the second size.

30. The method of claim 25, wherein the Nyquist region, being a function of Nyquist frequency, is symmetrically arranged with respect to a defined Nyquist frequency.

31. A computer program product comprising a computer readable medium having computer readable program code for implementing the method of claim 25.