# US 2010/0166274 A1 · Local motion compensation based on list mode data

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2010/0166274 A1
- Document ID: US20100166274A1
- Status: Published application
- Kind: A1
- Application number: 12/279,996
- Filed: 2007-02-05
- Published: 2010-07-01
- Earliest claimed: 2006-02-28
- Assignees: Philips Intellectual Property & Standards GmbH
- Inventors: Marc Busch; Ralph Brinks; Carsten Meyer
- Web page: https://patentgrep.com/patent/US20100166274A1

## CPC

- G06T 12/20: Inverse problem, i.e. transformations from projection space into object space
- A61B 6/037: {Emission tomography}
- G06T 2211/412: Dynamic
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- A61B: DIAGNOSIS; SURGERY; IDENTIFICATION (usefulness limited to only animals A61D)

## Abstract

Abstract: A method for locally correcting motion in an image reconstructed by a reconstruction system (42) of an imaging system (10) with raw data includes estimating a characteristic feature of a region of interest within the reconstructed image from the raw data, correcting the raw data associated with the region of interest for motion with the estimated region characteristic feature, and reconstructing a motion-corrected image corresponding to the region of interest with the corrected raw data.

## Claims

### Claim 1

A method for correcting motion in an image reconstructed from raw data, comprising:
estimating a characteristic feature of a region of interest within the reconstructed image from the raw data;
correcting the raw data associated with the region of interest for motion with the estimated region characteristic feature; and
reconstructing a motion-corrected image corresponding to the region of interest with the corrected raw data.

### Claim 2 (depends on 1)

The method as set forth in claim 1, wherein the characteristic feature is a center of mass.

### Claim 3 (depends on 1)

The method as set forth in claim 1, wherein the raw data is list mode data.

### Claim 4 (depends on 1)

The method as set forth in claim 1, wherein the region of interest is less than a field of view and localized to anatomy of interest.

### Claim 5 (depends on 1)

The method as set forth in claim 1, further including:
replacing the uncorrected region of interest within the image with the motion-corrected region of interest.

### Claim 6 (depends on 1)

The method as set forth in claim 1, further including concurrently presenting the motion-corrected region of interest and the uncorrected image in separate displays.

### Claim 7 (depends on 2)

The method as set forth in claim 2, further including, in connection with the steps of estimating the center of mass and correcting the raw data:
generating a plurality of intermediate regions of interest images from segments of the raw data for a number of short time frames;
adjusting, for each of the plurality of intermediate regions of interest images, a voxel value for one or more voxels that reside inside a region of interest and that intersect a corresponding line of response;
computing the center of mass for each of the intermediate regions of interest images;
computing a difference between each center of mass and a selected center of mass; and
computing a motion correction translation vector based on the differences.

### Claim 8 (depends on 7)

The method as set forth in claim 7, further including:
removing voxels with values less than a predetermined percentage of a maximum region of interest value before the estimation of the center of mass to decrease the influence from coincidences outside the region of interest.

### Claim 9 (depends on 7)

The method as set forth in claim 7, further including:
correcting the raw data for motion by shifting lines of response in the original data by an amount given by the motion correction translation vector.

### Claim 10 (depends on 1)

The method as set forth in claim 1, further including, in connection with the steps of estimating the characteristic feature and correcting the raw data:
computing a vector that represents a center point of a shortest line segment connecting a pair of lines of response in the raw data.

### Claim 11 (depends on 10)

The method as set forth in claim 10, wherein the vector includes a collection of points for a given number of successive pairs of lines of response.

### Claim 12 (depends on 10)

The method as set forth in claim 10, wherein the vector is computed as a function of
C
→
=
P
→
0
+
b
·
p
→
0
-
d
2

n
→
,
wherein {right arrow over (P)}0 and {right arrow over (P)}1 are points on a pair of lines of response and {right arrow over (p)}0 and {right arrow over (p)}1 are the unit direction vectors for these lines of response, d is a length of the shortest line segment connecting the lines of response and is defined by d=|({right arrow over (P)}0−{right arrow over (P)}1)·{right arrow over (n)}|, wherein n is a unit vector pointing in the direction of the shortest connection and is defined by {right arrow over (n)}:={right arrow over (p)}0×{right arrow over (p)}1/|{right arrow over (p)}0×{right arrow over (p)}1|, and b is defined by
b
=
[
P
0
y
-
P
1
y
+
dn
y
+
(
P
1
x
-
P
0
x
-
dn
x
)

p
0
y
p
0
x
]
/
(
p
1
y
-
p
1
x

p
0
y
p
0
x
)
,
wherein superscripts x and y denote the x and y components of the corresponding vector.

### Claim 13 (depends on 10)

The method as set forth in claim 10, further including computing a center of mass of the distribution in the region of interest.

### Claim 14 (depends on 13)

The method as set forth in claim 13, further including:
correcting the raw data for motion by shifting lines of response in the original data in accordance with the center of mass movement.

### Claim 15 (depends on 10)

The method as set forth in claim 10, further including refining the result by ignoring data outside of the region of interest.

### Claim 16 (depends on 1)

The method as set forth in claim 1, wherein correcting the raw data includes:
dividing the raw data into groups;
determining a location of the characteristic feature in each group from data in each group;
determining a deviation between the location of the characteristic feature in each group and the estimated characteristic feature;
adjusting the raw data in accordance with the determined deviations.

### Claim 17 (depends on 16)

The method as set forth in claim 16, wherein the raw data includes a plurality of lines of response (LORs) and the adjusting step includes:
translating the lines of response in accordance with the determined deviations to align the characteristic features of the groups.

### Claim 18 (depends on 1)

The method as set forth in claim 1, further including at least one of the following to improve the motion correction:
using time-of-flight information to facilitate estimating motion;
adjusting a temporal resolution to affect the motion estimation;
adjusting a number of events used for the estimation of the characteristic; and
employing an algorithm that determines an optimal setting for one or more parameters.

### Claim 19 (depends on 1)

A computer component programmed to execute the method of claim 1.

### Claim 20 (depends on 1)

A computer-readable medium that stores instructions for implementing the method of claim 1.

### Claim 21 (depends on 1)

An image corrected by the method of claim 1.

### Claim 22

A medical imaging system that corrects motion in a reconstructed image with data used to reconstruct the image, comprising:
a plurality of detectors that detect electron-positron annihilation events;
a coincident detector that identifies pairs of substantially simultaneous coincident events within the electron-positron annihilation events;
a line of response processor that processes the events and identifies lines of responses that connect events within each coincident event pair; and
one or more processors programmed to:
estimate a characteristic feature for the region of interest from the lines of response,
motion-correct the lines of response corresponding to the region of interest, and
reconstruct a motion corrected image corresponding to the region of interest from the motion corrected lines of response.

### Claim 23 (depends on 22)

The medical imaging system as set forth in claim 22, wherein the lines of response are list mode data.

### Claim 24 (depends on 22)

The medical imaging system as set forth in claim 22, wherein the region of interest is localized to less than a field of view of the medical imaging system.

### Claim 25 (depends on 22)

The medical imaging system as set forth in claim 22, wherein the processing components execute one of a simplified Siddon based algorithm and a vector based algorithm facilitate estimating a center of mass for the region of interest.

### Claim 26 (depends on 22)

The medical imaging system as set forth in claim 22, further including:
presentation components which present at least one of:
the motion corrected image,
concurrent displays of the motion corrected image and a non-corrected image, and
a non-corrected image with the motion corrected region of interest substituted for the corresponding region of interest of the non-corrected image.

### Claim 27 (depends on 22)

The medical imaging system as set forth in claim 22,
wherein the one or more processors are further programmed to:
groups the lines of response into temporally contiguous groups;
estimate a location of the characteristic feature in each group from the lines of response in each group;
shift the lines of response such that the locations of the characteristic feature in each group are aligned to generate the motion corrected lines of response for reconstructing.

### Claim 28

A computer implemented program for performing the steps of:
generating a plurality of intermediate regions of interest images from an motion-blurred region of interest for a number of short time frames;
adjusting, for each of the plurality of intermediate regions of interest images, a voxel value for one or more voxels that reside inside a region of interest and that intersect a corresponding line of response;
computing a center of mass for each of the intermediate regions of interest images;
combining the individual center of masses to compute an aggregate center of mass for a full time range of an acquisition;
computing a difference between each center of mass and the aggregated center of mass;
computing a motion correction translation vector based on the differences;
correcting data corresponding to the region of interest for motion by shifting lines of response in the data by an amount given by the motion correction translation vector; and
reconstructing a motion-corrected image corresponding to the region of interest with the corrected data.

### Claim 29 (depends on 28)

The computer implemented program as set forth in claim 28, wherein the region of interest is localized to anatomy of interest.

### Claim 30 (depends on 28)

The computer implemented program as set forth in claim 28, wherein the data corresponding to the region of interest is list mode data.

### Claim 31

A computer implemented program for performing the steps of:
computing a vector that represents a center point of a shortest line segment connecting a pair of lines of response;
computing the temporal motion of the center of mass from a distribution of center points in the region of interest;
correcting the data for motion by shifting lines of response in the original data to the overall center of mass; and
reconstructing a motion-corrected image corresponding to the region of interest with the corrected data.

### Claim 32

A method of forming a motion corrected image, comprising:
defining a region of interest within an uncorrected image; and
correcting the region of interest for differences caused by local motion by calculating a translation vector based on changes in a center of mass of the region of interest.

### Claim 33 (depends on 32)

The method as set forth in claim 32, further including:
generating a plurality of intermediate images from the region of interest for a number of short time frames;
adjusting, for each of the plurality of intermediate images, a voxel value for one or more voxels that reside inside the region of interest and that intersect a corresponding line of response;
computing a center of mass for each of the intermediate images; and
combining the individual center of masses to compute the center of mass of the region of interest for a full time range of an acquisition.

### Claim 34 (depends on 33)

The method as set forth in claim 33, further including:
computing a difference between the individual center of masses and the center of mass of the region of interest; and
computing the translation vector based on the differences.

### Claim 35 (depends on 32)

The method as set forth in claim 32, wherein the step of correcting the region of interest includes;
calculating the translation vector from list mode data corresponding to the region of interest.

### Claim 36 (depends on 32)

The method as set forth in claim 32, wherein the region of interest is localized to anatomy within a field of view.

### Claim 37

A method for locally correcting motion in an image, comprising:
localizing a region of interest within an uncorrected image;
correcting the localized region of interest for motion based on a characteristic feature of the localized region of interest; and
forming a corrected image by replacing the localized uncorrected region of interest with the localized corrected region of interest.

### Claim 38 (depends on 37)

The method as set forth in claim 37, wherein the characteristic feature is a center of mass.

### Claim 39 (depends on 37)

The method as set forth in claim 37, wherein the localized region of interest corresponds to list mode data.