Granted Patent B2

US 6,917,826 · App. 10/061,986 · Granted Jul 12, 2005

Abstract

Claims (73)

1. A method for automatic quantitative assessment of myocardial image quality, the method comprising:

providing a plurality of images;

detecting the heart position in each provided image;

segmenting non-heart activity relative to the detected heart position in each image;

measuring the liver-heart effect in each image relative to the segmentations;

fusing the measurements obtained in each image;

determining a quality number for the plurality of images responsive to the fused measurements; and

alerting a user to low image quality if the quality number is less than a threshold.

2. A method as defined in claim 1 , further comprising:

comparing the quality number with a threshold; and

performing appropriate actions if the quality number is poor relative to the threshold.

3. A method as defined in claim 1 wherein detecting the heart position in each image comprises:

receiving an image sequence;

generating a region of interest comprising the heart by 1-D pseudo-motion analysis of the image sequence;

localizing the heart center by a training-based algorithm;

fitting a curve to the localized heart centers of each image in the sequence;

refining the heart center positions of localized heart centers falling outside of a tolerance distance from the fitted curve; and

determining a bounding circle comprising the heart in accordance with the localized and refined heart center positions.

4. A method as defined in claim 3 wherein said curve is a sine curve for compensating for the motion of an imaging device.

5. A method as defined in claim 1 wherein said images comprise single-photon emission computed tomography images.

6. A system ( 100 ) for automatic quantitative assessment of myocardial image quality, the system comprising:

a heart detection unit ( 180 ) for heart detection, outlier detection and outlier correction;

a liver-heart artifact quantification unit ( 170 ) responsive to the heart detection unit for determining a quality number indicative of the severity of image-quality induced liver-heart; and

an output device ( 116 . 132 ) for alerting a user to low image quality if the quality number is less than a threshold.

7. A system ( 100 ) as defined in claim 6 wherein the images comprise single-photon emission computed tomography images.

8. A system ( 100 ) for automatic quantitative assessment of myocardial image quality, the system comprising:

a heart detection unit ( 180 ) for heart detection, outlier detection and outlier

a liver-heart artifact quantification unit ( 170 ) responsive to the heart detection unit for determining a quality number indicative of the severity of image-quality induced liver-heart artifacts; and

a CPU ( 102 ) in signal communication with said liver-heart artifact quantification unit ( 170 ) for processing the images.

9. A system ( 100 ) as defined in claim 8 , further comprising:

a display adapter ( 110 ) in signal communication with the CPU ( 102 ) for displaying the images; and

an I/O adapter ( 112 ) in signal communication with the CPU ( 102 ) for recalling the quantified features of the corresponding structure of the images to provide an automatic assessment of myocardial image quality corresponding to a selected image.

10. A system ( 100 ) as defined in claim 8 , further comprising:

a user interface adapter ( 114 ) in signal communication with the CPU ( 102 ) for at least receiving a selection decision for at least one image from a user.

11. A system for automatic quantitative assessment of myocardial image quality, the system comprising:

input means for providing a plurality of images;

detection means in signal communication with said input means for detecting the heart position in each provided image;

segmentation means in signal communication with said detection means for segmenting non-heart activity relative to a detected heart position in each image;

measurement means in signal communication with said segmentation means for measuring the liver-heart effect in each image based on the segmentations;

fusion means in signal communication with said measurement means for fusing measurements obtained in each image;

determination means in signal communication with said fusion means for determining a quality number for the plurality of images responsive to fused measurements; and

output means for alerting a user to low image quality if the quality number is less than a threshold.

12. A system as defined in claim 11 , further comprising:

comparison means in signal communication with said determination means for comparing the quality number with a threshold; and

action means in signal communication with said comparison means for performing appropriate actions if the quality number is poor relative to the threshold.

13. A system as defined in claim 11 wherein said detection means comprises:

reception means for receiving an image sequence;

generation means in signal communication with said reception means for generating a region of interest comprising the heart by 1-D pseudo-motion analysis of the image sequence;

localization means in signal communication with said generation means for localizing the heart center by a training-based algorithm;

fitting means in signal communication with said localization means for fitting a curve to the localized heart centers of each image in the sequence;

refining means in signal communication with said fitting means for refining the heart center positions of localized heart centers falling outside of a tolerance distance from the fitted curve; and

bounding means in signal communication with said refining means for determining a bounding circle comprising the heart in accordance with the localized and refined heart center positions.

14. A system as defined in claim 13 wherein said curve is a sine curve for compensating for the motion of an imaging device.

15. A system as defined in claim 11 wherein said images comprise single-photon emission computed tomography images.

16. A program storage device readable by machine, tangibly embodying a program of instructions executable by the machine to perform method steps for automatic quantitative assessment of myocardial image quality, the method steps comprising:

providing a plurality of images;

detecting the heart position in each provided image;

segmenting non-heart activity relative to the detected heart position in each image;

measuring the liver-heart effect in each image relative to the segmentations;

fusing the measurements obtained in each image; and

determining a quality number for the plurality of images responsive to the fused measurements.

17. A program storage device as defined in claim 16 , the method steps further comprising:

comparing the quality number with a threshold; and

performing appropriate actions if the quality number is poor relative to the threshold.

18. A program storage device as defined in claim 16 wherein the step of detecting the heart position in each image comprises:

receiving an image sequence;

generating a region of interest comprising the heart by 1-D pseudo-motion analysis of the image sequence;

localizing the heart center by a training-based algorithm;

fitting a curve to the localized heart centers of each image in the sequence;

refining the heart center positions of localized heart centers falling outside of a tolerance distance from the fitted curve; and

determining a bounding circle comprising the heart in accordance with the localized and refined heart center positions.

19. A program storage device as defined in claim 18 wherein said curve is a sine curve for describing the motion of an imaging device.

20. A program storage device as defined in claim 16 wherein said images comprise single-photon emission computed tomography images.

Assignments (6)