Method, apparatus, and product for accurately determining the intima-media thickness of a blood vessel

• US 6,817,982 B2
• Filed: 04/07/2003
• Issued: 11/16/2004
• Est. Priority Date: 04/19/2002
• Status: Active Grant

• Current Assignee: FUJIFILM SonoSite, Inc. (Fujifilm Holdings Corporation)
• Original Assignee: Sonosite Incorporated (Fujifilm Holdings Corporation)
• Inventors: Fritz, Helmuth, Fritz, Terry
• Primary Examiner(s): Jaworski, Francis J.
• Application Number:
  US10/407,682
• Publication Number:
  US 20030199762A1
• Time in Patent Office: 589 Days
• Field of Search: 600/437, 600/443, 600/447, 600/463, 600/466-467, 328/100, 328/128, 345/215, 345/419, 345/660, 345/664-668
• US Class Current: 600/443
• CPC Class Codes: A61B 8/0858 involving measuring tissue layers, e.g. skin, interfaces
• CPC Class Codes: G06T 2207/30101 Blood vessel; Artery; Vein; Vascular
• CPC Class Codes: G06T 7/0012 Biomedical image inspection
• CPC Class Codes: G06T 7/60 Analysis of geometric attributes

Abstract

An apparatus, automated method technique and product for accurately determining the intima-media thickness of a blood vessel.

Claims

1. A method for determining the intima-media thickness of the near and/or far wall of a blood vessel comprising:

a. capturing pixel images from an ultrasound machine'"'"'s scan of a desired segment of blood vessels and inputting said blood vessel pixel images into a computer,

b. inputting an analyzer program into the computer which determines the calibrated pixel size of the ultrasound image using at least one of the following mechanisms;

i. operator specifies two points on the ultrasound image and also specifies the distance between the two specified points, after which the analyzer program computes the calibrated pixel size and assumes this size for both vertical and horizontal dimensions,

ii. operator specifies two horizontal points as well as the distance between these two horizontal points and also specifies two vertical points and the distance between these two vertical points, after which the analyzer program computes the horizontal calibrated pixel size and the vertical calibrated pixel size,

iii. the analyzer program recognizes known landmarks and characteristics on the ultrasound image and determines the image from a known ultrasound device with a known zoom/scale factor, and a known distance between found landmarks, to determine how many pixels are between a set of vertical and/or horizontal landmarks without operator intervention,

c. activating the analyzer program to computer locate via pixel image intensities the lumen/intima and media/adventitia boundaries of a blood vessel'"'"'s near and/or far wall using at least one of the following operator inputs;

i. operator specifies a single pixel image position as the analyzer program'"'"'s horizontal center, left, or right position in the measurement region,

ii. operator specifies a left, right, and/or overall width constraint

iii. operator specifies an outer constraint distance, boundary line, multi-segment boundary line, boundary spline curve or polynomial boundary curve to constrain the media/adventitia pixel path to within this boundaries distance from the center of the blood vessel,

iv. operator specifies an inner constraint distance, boundary line, multi-segment boundary line, boundary spline curve or polynomial boundary curve to constrain the media/adventitia pixel path to no closer than this boundaries distance to the center of the blood vessel,

v. operator specifies an inner constraint distance, boundary line, multi-segment boundary line, boundary spline curve or polynomial boundary curve to constrain the lumen/intima pixel path to no closer than this boundaries distance to the center of the blood vessel,

vi. operator specifies an outer constraint distance, boundary line, multi-segment boundary line, boundary spline curve or polynomial boundary curve to constrain the lumen/intima pixel path to within this boundaries distance from the center of the blood vessel,

vii. operator specifies a smoothness factor to determine how jaggy the lumen/intima and/or the media/adventitia pixel paths can be, and

d. computer calculating the averaged distance between lumen/intima and media/adventitia boundaries.

2. A method for determining the intima-media thickness of the near and/or far wall of a blood vessel according to claim 1, wherein the analyzer program optionally computes an image intensity histogram covering the entire measurement region which is then used to determine the dynamic range of pixel intensities for the purposes of determining typical lumen pixel value and typical adventitial pixel value for reference points to be used when finding lumen pixel path and adventitial pixel path.

3. A method for determining the intima-media thickness of the near and/or far wall of a blood vessel according to claim 1, wherein the computer program analyzer uses slope correction to compensate for images of blood vessels that are sloping.

4. A method for determining the intima-media thickness of the near and/or far wall of a blood vessel according to claim 1, wherein the computer program analyzer compensates for tapering of the diameter of the vessel as a function of distance from a point of bifurcation splitting from a single vessel into two vessels.

5. A method for determining the intima-media thickness of the near and/or far wall of a blood vessel according to claim 1, including computer generating a report via an output device stratifying the patient'"'"'s general atherosclerotic burden and/or comparing the patient'"'"'s IMT value with the general population.

6. A method for determining the intima-media thickness of the near and/or far wall of a blood vessel according to claim 1, including inputting into the computer a database of the intima-media thickness, and possibly the presence of plaque combined with patient information consisting of some or all, but not limited to, age, gender, typical cardiovascular risk factors, and wherein the computer program includes calculations to allow the supplementation and/or augmentation into a general population data pool as well as to allow patient follow up on how a patient'"'"'s IMT varies over time.

7. A method for determining the intima-media thickness of the near and/or far wall of a blood vessel according to claims 1, including employing any or all of the following to address operator dependent factors that reduce accuracy:

a. minimizing gain between the extremes of high and low of the ultrasound instrument TGC and overall power gain settings,

b. compensating for IMT as a function of distance from dilation using the CCA dilation landmark position to the measured IMT segment to account for typical variations in IMT thickness as a function of distance from the dilation point,

c. compensating for the probe angle positions in relation to the neck surface during image acquisition because of the anatomically eccentric variation in IMT thickening,

d. factoring for the head position during image acquisition,

e. ensuring inclusion of near wall measurements for true atherosclerotic burden stratification,

f. factoring for ECG synchrony, which varies within each cardiac cycle, and

g. compensating for the cross-sectioning of the maximal CCA diameter, when measuring the IMT from an image captured from a cross section view of the vessel that does not include the maximal diameter of the vessel.

8. A method for determining the intima-media thickness of the near and/or far wall of a blood vessel according to claim 1, including manually assisting the automated algorithm in varying steps for images with poor contrast or signal-to-noise ratio where the automated values are obviously in error, by the operator manually assisting in any or all of the following:

a) locating the adventitia,

b) locating the lumen,

c) determining the curvature/slope of the intima/media layer,

d) constraining the search region for the lumen/intima boundary,

e) constraining the search region for the media/adventitia boundary,

f) increasing, decreasing or laterally moving the region of measurement to find a less noisy region of the intima/media layer, and

g) strengthening the aggressiveness (more boldness) with which the various “

horizontal”

pathways parallel to vessel wall can continue through even lower contrast regions.

9. A method for determining the intima-media thickness of the near and/or far wall of a blood vessel according to claim 1, including computing the intensity histogram in the region of the measured IMT and using this information as a confidence indicator to the operator as to how good the automated algorithm is performing, and if the confidence is too low, the operator manually assists in the IMT measurement or ignores this particular image altogether.

10. A method for determining the intima-media thickness of the near and/or far wall of a blood vessel according to claim 9, wherein the mechanisms that use histogram information as a confidence indicator could include one or more of the following:

a) displaying the histogram to the operator for comparison with a good quality or poor quality histogram,

b) generating a numerical correlation between the computed histogram and a known good quality shaped histogram for ultrasound images that the automated measurement system is known to perform well against and displaying the degree to which this computed histogram correlates using one or more of the following display mechanisms;

i. a color coded indicator that varies in colors from a low confidence color like red, to a high confidence color like green,

ii. a numerical display of confidence as a percentage, where a low percentage would indicate a poor confidence and a high percentage would indicate a high confidence, and

iii. status indicator in the form of a “

meter display”

that displays the confidence on a range from poor to excellent; and

c) Overlaying the computed histogram on top of known baseline good histograms and/or poor histograms to visually correlate how close the computed histogram matches baseline histograms.

11. An apparatus for determining the intima-media thickness of the near and/or far wall of a blood vessel comprising:

a. means for capturing pixel images from an ultrasound machine'"'"'s scan of a desired segment of blood vessels and inputting said blood vessel pixel images into

b. a standalone or embedded computer,

c. an analyzer program inputted into the standalone or embedded computer which determines the calibrated pixel size of the ultrasound image using at least one of the following mechanisms;

i. operator specifies two points on the ultrasound image and also specifies the distance between the two specified points, after which the analyzer program computes the calibrated pixel size and assumes this size for both vertical and horizontal dimensions,

ii. operator specifies two horizontal points as well as the distance between these two horizontal points and also specifies two vertical points and the distance between these two vertical points, after which the analyzer program computes the horizontal calibrated pixel size and the vertical calibrated pixel size,

iii. the analyzer program recognizes known landmarks and characteristics on the ultrasound image and determines the image is from a known ultrasound device and is also at a known zoom/scale factor and can therefore determine how many pixels are between a set of vertical and/or horizontal landmarks and, since it knows from which ultrasound device the image originated, it knows the distance between the found landmarks, and can therefore computer the calibrated pixel size without operator intervention since it has determined the number of pixels and knows the distance across those pixels,

whereby the analyzer program causes the computer to locate via pixel image intensities the lumen/intima and media/adventitia boundaries of a blood vessel'"'"'s near and/or far wall using at least one of the following operator inputs;

i. operator specifies a single pixel image position as the analyzer program'"'"'s horizontal center, left, or right position in the measurement region,

ii. operator specifies a left, right, and/or overall width constraint

iii. operator specifies an outer constraint distance, boundary line, multi-segment boundary line, boundary spline curve or polynomial boundary curve to constrain the media/adventitia pixel path to within this boundaries distance from the center of the blood vessel,

iv. operator specifies an inner constraint distance, boundary line, multi-segment boundary line, boundary spline curve or polynomial boundary curve to constrain the media/adventitia pixel path to no closer than this boundaries distance to the center of the blood vessel,

v. operator specifies an inner constraint distance, boundary line, multi-segment boundary line, boundary spline curve or polynomial boundary curve to constrain the lumen/intima pixel path to no closer than this boundaries distance to the center of the blood vessel,

vi. operator specifies an outer constraint distance, boundary line, multi-segment boundary line, boundary spline curve or polynomial boundary curve to constrain the lumen/intima pixel path to within this boundaries distance from the center of the blood vessel,

vii. operator specifies a smoothness factor to determine how jaggy the lumen/intima and/or the media/adventitia pixel paths can be,

and calculates the averaged distance between lumen/intima and media/adventitia boundaries.

12. An apparatus for determining the intima-media thickness of the near and/or far wall of a blood vessel according to claim 11, wherein the computer program uses slope correction to compensate for images of blood vessels that are sloping.

13. An apparatus for determining the intima-media thickness of the near and/or far wall of a blood vessel according to claim 11, wherein the computer program compensates for tapering of the diameter of the vessel as a function of distance from a point of bifurcation (splitting from a single vessel into two vessels).

14. An apparatus for determining the intima-media thickness of the near and/or far wall of a blood vessel according to claim 11, including computer generating means to generate a report via an output device stratifying the patient'"'"'s general atherosclerotic burden and/or comparing the patient'"'"'s IMT value with the general population.

15. An apparatus for determining the intima-media thickness of the near and/or far wall of a blood vessel according to claim 11, including a computer database wherein the intima-media thickness, and possibly the presence of plaque combined with patient information consisting of some or all, but not limited to, age, gender, typical cardiovascular risk factors is added to allow the supplementation and/or augmentation to the general population data pool as well as to allow patient follow up on how his/her IMT varies over time.

16. An apparatus that determines the intima-media thickness of the near and/or far wall of a blood vessel according to claim 11, wherein the computer program optionally allows manually assisting the automated algorithm in varying steps for images with poor contrast or signal-to-noise ratio where the automated values are obviously in error, by the operator manually assisting in any or all of the following:

a) locating the adventitia,

c) locating the lumen,

c) determining the curvature/slope of the intima/media layer,

d) constraining the search region for the lumen/intima boundary,

e) constraining the search region for the media/adventitia boundary,

f) increasing, decreasing or laterally moving the region of measurement to find a less noisy region of the intima/media layer, and

g) strengthening the aggressiveness (more boldness) with which the various “

horizontal”

pathways parallel to vessel wall can continue through even lower contrast regions.

17. An apparatus that determines the intima-media thickness of the near and/or far wall of a blood vessel according to claim 11, wherein the computer program computes the intensity histogram in the region of the measured IMT and uses this information as a confidence indicator to the operator as to how good the automated algorithm is performing, and if the confidence is too low, the operator manually assists in the IMT measurement or ignores this particular image altogether.

18. An apparatus that determines the intima-media thickness of the near and/or far wall of a blood vessel according to claim 17, including program mechanisms that use histogram information as a confidence indicator like one or more of the following:

a) displaying the histogram to the operator for comparison with a good quality or poor quality histogram,

b) generating a numerical correlation between the computed histogram and a known good quality shaped histogram for ultrasound images that the automated measurement system is known to perform well against and displaying the degree to which this computed histogram correlates using one or more of the following display mechanisms;

i. a color coded indicator that varies in colors from a low confidence color like red, to a high confidence color like green,

ii. a numerical display of confidence as a percentage, where a low percentage would indicate a poor confidence and a high percentage would indicate a high confidence, and

iii. status indicator in the form of a “

meter display”

that displays the confidence on a range from poor to excellent; and

c) overlaying the computed histogram on top of known baseline good histograms and/or poor histograms to visually correlate how close the computed histogram matches baseline histograms.

19. An apparatus for determining the intima-media thickness of the near and/or far wall of a blood vessel according to claim 11, including a conventional global communication network to transmit off-site measurements to a remote location for digital analysis via a modem, network card, blue tooth interface or any other interface hook-ups to any known data transmission networks such as terrestrial and wireless phone networks, optical data transmission networks, local area networks (LAN), wide area networks (WAN) and all other known and unknown transmission networks and mediums to access a central computer.

20. A computer program product embedded on a computer-readable medium for determining the intima-media thickness of the near and/or far wall of a blood vessel comprising:

a. capturing pixel images from an ultrasound machine'"'"'s scan of a desired segment of blood vessels and inputting said blood vessel pixel images into a standalone or embedded computer,

b. determining the calibrated pixel size of the ultrasound image using at least one of the following mechanisms;

i. operator specifies two points on the ultrasound image and also specifies the distance between the two specified points, after which the analyzer program computes the calibrated pixel size and assumes this size for both vertical and horizontal dimensions,

ii. operator specifies two horizontal points as well as the distance between these two horizontal points and also specifies two vertical points and the distance between these two vertical points, after which the analyzer program computes the horizontal calibrated pixel size and the vertical calibrated pixel size,

iii. the analyzer program recognizes known landmarks and characteristics on the ultrasound image and determines the image is from a known ultrasound device and is also at a known zoom/scale factor and can therefore determine how many pixels are between a set of vertical and/or horizontal landmarks and, since it knows from which ultrasound device the image originated, it knows the distance between the found landmarks, and can therefore computer the calibrated pixel size without operator intervention since it has determined the number of pixels and knows the distance across those pixels,

c. locating via pixel image intensities the lumen/intima and media/adventitia boundaries of a blood vessel'"'"'s near and/or far wall using one or more of the following operator inputs;

i. operator specifies a single pixel image position as the analyzer program'"'"'s horizontal center, left, or right position in the measurement region,

ii. operator specifies a left, right, and/or overall width constraint

iii. operator specifies an outer constraint distance, boundary line, multi-segment boundary line, boundary spline curve or polynomial boundary curve to constrain the media/adventitia pixel path to within this boundaries distance from the center of the blood vessel,

iv. operator specifies an inner constraint distance, boundary line, multi-segment boundary line, boundary spline curve or polynomial boundary curve to constrain the media/adventitia pixel path to no closer than this boundaries distance to the center of the blood vessel,

v. operator specifies an inner constraint distance, boundary line, multi-segment boundary line, boundary spline curve or polynomial boundary curve to constrain the lumen/intima pixel path to no closer than this boundaries distance to the center of the blood vessel,

vi. operator specifies an outer constraint distance, boundary line, multi-segment boundary line, boundary spline curve or polynomial boundary curve to constrain the lumen/intima pixel path to within this boundaries distance from the center of the blood vessel,

vii. operator specifies a smoothness factor to determine how jaggy the lumen/intima and/or the media/adventitia pixel paths can be,

d. calculating the averaged distance between lumen/intima and media/adventitia boundaries.

21. A computer program product embedded on a computer-readable medium for determining the intima-media thickness of a blood vessel according to claim 20, wherein the analyzer program sequence optionally computes an image intensity histogram covering the entire measurement region which is then used to determine the dynamic range of pixel intensities for the purposes of determining typical lumen pixel value and typical adventitial pixel value for reference points to be used when finding lumen pixel path and adventitial pixel path.

22. A computer program product embedded on a computer-readable medium for determining the intima-media thickness of a blood vessel according to claim 20, wherein the computer program analyzer sequence uses slope correction to compensate for images of blood vessels that are sloping.

23. A computer program product embedded on a computer-readable medium for determining the intima-media thickness of a blood vessel according to claim 20, wherein the computer program analyzer sequence compensates for tapering of the diameter of the vessel as a function of distance from a point of bifurcation splitting from a single vessel into two vessels.

24. A computer program product embedded on a computer-readable medium for determining the intima-media thickness of a blood vessel according to claim 20, wherein the computer program analyzer sequence generates a report via an output device stratifying the patient'"'"'s general atherosclerotic burden and/or comparing the patient'"'"'s IMT value with the general population.

25. A computer program product embedded on a computer-readable medium for determining the intima-media thickness of a blood vessel according to claim 20, wherein the computer program analyzer sequence operates on a database inputted into the computer of the intima-media thickness, and possibly the presence of plaque combined with patient information consisting of some or all, but not limited to, age, gender, typical cardiovascular risk factors, to calculate and allow the supplementation and/or augmentation into a general population data pool as well as to allow patient follow up on how a patient'"'"'s IMT varies over time.

26. A computer program product embedded on a computer-readable medium for determining the intima-media thickness of a blood vessel according to claim 20, wherein the computer program analyzer sequence allows the employment of any or all of the following to address operator dependent factors that reduce accuracy:

a. minimizing gain between the extremes of high and low of the ultrasound instrument TGC and overall power gain settings,

b. compensating for IMT as a function of distance from dilation using the CCA dilation landmark position to the measured IMT segment to account for typical variations in IMT thickness as a function of distance from the dilation point,

c. compensating for the probe angle positions in relation to the neck surface during image acquisition because of the anatomically eccentric variation in IMT thickening,

d. factoring for the head position during image acquisition,

e. ensuring inclusion of near wall measurements for true atherosclerotic burden stratification,

f. factoring for ECG synchrony, which varies within each cardiac cycle, and

g. compensating for the cross-sectioning of the maximal CCA diameter, when measuring the IMT from an image captured from a cross section view of the vessel that does not include the maximal diameter of the vessel.

27. A computer program product embedded on a computer-readable medium for determining the intima-media thickness of a blood vessel according to claim 20, wherein the computer program analyzer sequence allows manually assisting the automated algorithm in varying steps for images with poor contrast or signal-to-noise ratio where the automated values are obviously in error, by the operator manually assisting in any or all of the following:

a) locating the adventitia,

d) locating the lumen,

c) determining the curvature/slope of the intima/media layer,

d) constraining the search region for the lumen/intima boundary,

e) constraining the search region for the media/adventitia boundary,

f) increasing, decreasing or laterally moving the region of measurement to find a less noisy region of the intima/media layer, and

g) strengthening the aggressiveness (more boldness) with which the various “

horizontal”

pathways parallel to vessel wall can continue through even lower contrast regions.

28. A computer program product embedded on a computer-readable medium for determining the intima-media thickness of a blood vessel according to claim 20, including computing the intensity histogram in the region of the measured IMT and using this information as a confidence indicator to the operator as to how good the automated algorithm is performing, and if the confidence is too low, the operator manually assists in the IMT measurement or ignores this particular image altogether.

29. A computer program product embedded on a computer-readable medium for determining the intima-media thickness of a blood vessel according to claim 20, wherein the mechanisms that use histogram information as a confidence indicator could include one or more of the following:

a) displaying the histogram to the operator for comparison with a good quality or poor quality histogram,

b) generating a numerical correlation between the computed histogram and a known good quality shaped histogram for ultrasound images that the automated measurement system is known to perform well against and displaying the degree to which this computed histogram correlates using one or more of the following display mechanisms;

iv. a color coded indicator that varies in colors from a low confidence color like red, to a high confidence color like green,

v. a numerical display of confidence as a percentage, where a low percentage would indicate a poor confidence and a high percentage would indicate a high confidence, and

vi. status indicator in the form of a “

meter display”

that displays the confidence on a range from poor to excellent; and

c) Overlaying the computed histogram on top of known baseline good histograms and/or poor histograms to visually correlate how close the computed histogram matches baseline histograms.