System and method for 3-D medical imaging using 2-D scan data

• US 5,582,173 A
• Filed: 09/18/1995
• Issued: 12/10/1996
• Est. Priority Date: 09/18/1995
• Status: Expired due to Term

• Current Assignee: Siemens Medical Solutions USA Incorporated (Siemens AG)
• Original Assignee: Siemens Medical Systems, Inc. (Siemens AG)
• Inventors: Li, Ming
• Primary Examiner(s): MANUEL, GEORGE C
• Application Number:
  US08/529,778
• Time in Patent Office: 449 Days
• Field of Search: 128/660.06, 128/660.07, 128/660.08, 128/660.09, 128/660.10, 128/661.01, 128/661.10, 128/916, 128/660.04, 73/620, 73/625
• US Class Current: 600/443
• CPC Class Codes: G01S 15/8934 using a dynamic transducer configuration
• CPC Class Codes: G01S 15/8993 Three dimensional imaging systems
• CPC Class Codes: G01S 7/52034 Data rate converters
• CPC Class Codes: G01S 7/52046 Techniques for image enhancement involving transmitter or receiver image enhancement by image data processing G06T5/00
• CPC Class Codes: Y10S 128/916 Ultrasound 3-D imaging

Abstract

A region of tissue is scanned as a series of 2-D frames. The correlation of substantially homogeneous speckle regions in the frames is determined and is compared with a precalibrated speckle correlation-versus-distance function to give an estimate of the distance between the frames. In some applications, the frames are partitioned into sub-frames, whose correlation values and distances are combined according to predetermined functions to give a distance estimate between the frames as a whole. The reliability of the distance estimates is improved and evaluated in various ways, for example, by comparing combinations of possible distances from end frames via intermediate frames, and by comparing computed frame or sub-frame velocities with a known or estimated transducer velocity. Once the relative distances of the 2-D frames are estimated, a 3-D image is compiled using registration techniques. Image frames need not be parallel.

Claims

1. A method for generating a three-dimensional representation of a region of tissue scanned by a transducer, comprising the following steps:

A) pre-determining a calibrated speckle correlation-versus-distance function for the transducer to be used, for the tissue to be scanned, or both;

B) generating a series of 2-D image frames, divided into image elements;

C) identifying image elements representing speckle;

D) for each scanned image frame, estimating actual speckle correlation with at least one other frame based on at least the image elements in corresponding portions of the respective frames;

E) estimating the distance between pairs of image frames by evaluating the correlation-versus-distance function with the corresponding estimated actual speckle correlation for the pairs of frames, including estimating the distance between a reference and at least one other image frame;

F) for each frame for which correlation has been estimated, determining its relative displacement from the reference by evaluating the correlation-versus-distance function with the estimated actual speckle correlation;

G) registering the different frames in a three-dimensional format as a 3-D representation; and

H) displaying the 3-D representation.

2. A method as defined in claim 1, further including the following steps:

dividing each scanned image frame into a pattern of the sub-frames, each sub-frame comprising a pre-determined sub-set of the image elements in the scanned image frame and corresponding to one sub-frame in at least one other scanned image frame;

identifying certain ones of the sub-frames in each scanned image frame as speckle sub-frames; and

estimating the actual speckle correlation between pairs of scanned image frames as a predetermined function of actual speckle correlation between corresponding pairs of speckle sub-frames in the pairs of scanned image frames.

3. A method as defined in claim 2, further including following steps:

for each corresponding pair of sub-frames in predetermined pairs of scanned image frames, computing a sub-frame correlation value; and

determining a sub-frame distance between the predetermined pairs of scanned image frames as a predetermined function of the sub-frame correlation values of the image frames whose correlation values exceed a predetermined minimum correlation reliability value.

4. A method as defined in claim 2, further including the following steps:

generating a plurality of the 2-D image frames to be non-parallel, the displacement of each such non-parallel frame from the reference thereby having a rotational component; and

estimating angular displacement of adjacent non-parallel image frames as a predetermined function of the distance estimates of corresponding sub-frames of the adjacent non-parallel image frames.

5. A method as defined in claim 2, further including the following steps:

estimating a local transducer velocity over a plurality of image frames;

calculating a frame velocity for each of selected ones of the scanned image frames;

comparing each frame velocity with the estimated local transducer velocity; and

marking as a spurious frame any frame whose frame velocity differs by more than a predetermined velocity threshold function value from the local transducer velocity.

6. A method as defined in claim 5, further including the following steps:

calculating a velocity-based distance value for each spurious frame, wherein the velocity-based distance value is defined as the product of the local transducer velocity and a predetermined time value; and

for each spurious frame, setting its distance from the reference equal to a predetermined functional value of the corresponding velocity-based distance value.

7. A method as defined in claim 2, further including the following steps:

for each corresponding pair of sub-frames in predetermined pairs of scanned image frames, computing a sub-frame correlation value;

determining a sub-frame distance between the pairs of sub-frames by evaluating the correlation-versus-distance function with the sub-frame correlation values;

for each corresponding pair of speckle sub-frames in the pairs of scanned image frames, estimating a sub-frame transducer velocity value;

calculating a frame velocity;

comparing each sub-frame velocity with the frame velocity; and

estimating the relative displacement of the frames as a predetermined function of the sub-frame distances for only those sub-frames whose sub-frame transducer velocity values differ by less than predetermined functional reliability value from the frame velocity.

8. A method as defined in claim 1, further including the following steps:

dividing each scanned image frame into a pattern of the sub-frames, each sub-frame comprising a pre-determined sub-set of the image elements in the scanned image frame and corresponding to one sub-frame in at least one other scanned image frame;

for each sub-frame of each scanned image frame, estimating a sub-frame speckle probability, defined as the probability that the sub-frame represents a speckle region of the tissue;

for each corresponding pair of sub-frames in predetermined pairs of scanned image frames, computing a sub-frame correlation value; and

determining a distance between the predetermined pairs of scanned image frames as a predetermined function of the sub-frame correlation values of the image frames.

9. A method as defined in claim 8, further including the step of computing a weight for each sub-frame correlation value as a predetermined function of the speckle probability for the corresponding sub-frame;

in which;

the step of determining the distance comprises forming a weighted average of the sub-frame correlation values and evaluating the correlation-versus-distance function using the weighted average.

10. A method as defined in claim 2, further including the following steps:

for each corresponding pair of sub-frames in predetermined pairs of scanned image frames, computing a sub-frame correlation value;

computing a weight for each sub-frame correlation value as a predetermined function of the distance of each speckle sub-frame from a focal region of the transducer; and

determining the distance between the predetermined pairs of scanned image frames by forming a weighted average of the sub-frame correlation values and evaluating the correlation-versus-distance function using the weighted average.

11. A method as defined in claim 1, further including the following steps:

for selected pairs of image frames for which at least one intermediate image frame is located between them, designating the frames in the selected pairs as end frames;

selecting a plurality of combinations of the end frames and of the intermediate frames;

summing the estimated distances between all adjacent pairs of the end frames and the intermediate frames to form a possible total distance; and

estimating the distance between the end frames as a predetermined function of the possible total distances for the selected plurality of combinations.

12. A method as defined in claim 1, further including the following steps:

estimating a local transducer velocity over a plurality of image frames;

calculating a frame velocity for each of selected ones of the scanned image frames;

comparing each frame velocity with the estimated local transducer velocity; and

marking as a spurious frame any frame whose frame velocity differs by more than a predetermined velocity threshold function value from the local transducer velocity.

13. A method as defined in claim 12, further including the following steps:

calculating a velocity-based distance value for each spurious frame, wherein the velocity-based distance value is defined as the product of the local transducer velocity and a predetermined time value; and

for each spurious frame, setting its distance from the reference equal to a predetermined functional value of the corresponding velocity-based distance value.

14. A method as defined in claim 1, further including the following steps:

generating the 2-D image frames as planes extending in a depth direction; and

compensating the image frames for transducer motion in the depth direction before the step of registering the different frames in a three-dimensional format.

15. A method as defined in claim 1, further including the following steps:

carrying out steps A)-F) for a plurality of scans of the region, each scan extending in a different principal direction; and

registering the frames of all scans into a single 3-D registration.

16. A system for generating a three-dimensional representation of a region of tissue including:

A) an ultrasonic transducer scanning the region as a series of 2-D image frames, each frame comprising a pattern of image elements;

B) processing means;

1) for identifying in each 2-D image frame image elements representing speckle;

2) for each scanned image frame, for estimating actual speckle correlation with at least one other frame based on at least the image elements in corresponding portions of the respective frames;

3) for estimating the distance between pairs of image frames by evaluating a predetermined calibrated correlation-versus-distance function with the corresponding estimated actual speckle correlation for the pairs of frames, including estimating the distance between a reference and at least one other image frame;

4) for each frame for which correlation has been estimated, for determining its relative displacement from the reference by evaluating the correlation-versus-distance function with the estimated actual speckle correlation; and

5) for registering the different frames in a three-dimensional format as a 3-D representation;

C) a memory including a 2-D frame data portion for storing the 2-D image frames and a correlation data portion for storing predetermined parameters of the correlation-versus-distance function; and

D) a display means for displaying the 3-D representation.