Method for segmenting features in an image
First Claim
1. A method for analyzing registered images of an object, steps of which comprise:
- (a) acquiring X images of the object where X is a positive integer, greater than one and wherein each image is formed by a two dimensional array of image elements;
(b) creating a X-dimensional feature space histogram of intensity levels of the image elements in the X images;
(c) locating a first centroid in the feature space histogram which centroid corresponds to a first feature in the X images;
(d) locating a second centroid in the feature space histogram which centroid corresponds to a second feature in the X images;
(e) defining a Cartesian coordinate system in the feature space histogram, wherein one axis of the Cartesian coordinate system passes through the first and second centroids;
(f) decomposing a vector, that defines the location in the feature space histogram associated with a given image element, into a component vector along the one axis;
(g) determining from the component vector, a fractional quantity of the first feature in the given image element;
(h) repeating steps (f) through (g) for a plurality of image elements in one of the X images; and
(i) deriving a measurement of a physical parameter of the first feature from a plurality of fractional quantities. PG,25
1 Assignment
0 Petitions
Accused Products
Abstract
A dual echo magnetic resonance imaging system produces two registered images of a patient in which the images have different contrast relationships between different tissue types. A two dimensional feature space histogram of the two images is produced and a separate centroid is located in the feature space histogram for each one of a pair of tissue types. A Cartesian coordinate system is defined in the feature space so that one axis of the system passes through the two centroids. Vector decomposition is employed to project each image element data point in the feature space onto a point on the one axis. The fractional quantity of each tissue type present in the image element is determined based upon the Euclidean distances from that axis point to the respective centroids. The fractional quantity is calculated for each element in the original images to form a pair of tissue images. The elements of a tissue image are processed to measure the amount of that tissue type in the imaged portion of the patient.
-
Citations
12 Claims
-
1. A method for analyzing registered images of an object, steps of which comprise:
-
(a) acquiring X images of the object where X is a positive integer, greater than one and wherein each image is formed by a two dimensional array of image elements; (b) creating a X-dimensional feature space histogram of intensity levels of the image elements in the X images; (c) locating a first centroid in the feature space histogram which centroid corresponds to a first feature in the X images; (d) locating a second centroid in the feature space histogram which centroid corresponds to a second feature in the X images; (e) defining a Cartesian coordinate system in the feature space histogram, wherein one axis of the Cartesian coordinate system passes through the first and second centroids; (f) decomposing a vector, that defines the location in the feature space histogram associated with a given image element, into a component vector along the one axis; (g) determining from the component vector, a fractional quantity of the first feature in the given image element; (h) repeating steps (f) through (g) for a plurality of image elements in one of the X images; and (i) deriving a measurement of a physical parameter of the first feature from a plurality of fractional quantities. PG,25 - View Dependent Claims (3)
-
-
2. The method as recited in claim I wherein each element in the X images corresponds to an amount of space in the object;
- and further comprises;
separately multiplying the amount of space by each fractional quantity of the first feature determined at step (g) to produce a plurality of fractional measurements; and summing the plurality of fractional measurements to produce a total space measurement for the first feature.
- and further comprises;
-
4. A method for generating an image of an patient, steps of which comprise:
-
(a) transmitting electromagnetic radiation toward the patient; (b) detecting electromagnetic radiation carrying image information from the patient; (c) forming a first image and a second image from detected radiation wherein each of the first and second images is a two dimensional array of image elements; (d) creating a two dimensional feature space histogram of intensity levels of the image elements in the first and second images; (e) locating a first centroid for a first tissue type in the feature space histogram; (f) locating a second centroid for a second tissue type in the feature space histogram; (g) defining a Cartesian coordinate system in the feature space histogram wherein the Cartesian coordinate system has an origin at one of the first and second centroids and has one axis that passes through the first and second centroids; (h) for a given image element in the first image, projecting an associated point in the feature space histogram onto a point on the one axis; (i) determining from the point on the one axis a fractional quantity of the first type of tissue in the given image element; (j) setting a value of an element in a first tissue image to the fractional quantity of the first type of tissue; (k) repeating steps (h) through (j) for a plurality of image elements in the first image; and (l) displaying the first tissue image. - View Dependent Claims (5, 6, 7, 8, 9)
-
-
10. A method for generating an image of a patient, steps of which comprise:
-
(a) detecting electromagnetic radiation carrying image information from the patient; (b) forming a first image and a second image from detected radiation wherein each of the first and second images is a two dimensional array of image elements; (c) creating a two-dimensional feature space histogram of intensity levels of the image elements in the first and second images; (d) locating first, second and third centroids in the feature space histogram which centroids respectively correspond to first, second and third features in the first and second images; (e) defining a first Cartesian coordinate system in the feature space histogram wherein one axis of the first Cartesian coordinate system passes through the first and second centroids; (f) decomposing a vector that defines the location for a given image element in one of the plurality of images into a first component vector along the one axis; (g) determining from the first component vector, a first fractional amount of the first feature in the given image element, and a second fractional amount of the second feature in the given image element; (h) storing the first and second fractional amounts in a memory device; (i) repeating steps (f) through (h) for a plurality of image elements in the one image; and (j) defining a second Cartesian coordinate system in the feature space histogram wherein one axis of the second Cartesian coordinate system passes through the first and third centroids; (k) decomposing a vector that defines the location for a given image element in one of the plurality of images into a second component vector along the one axis of the second Cartesian coordinate system; (l) determining from the second component vector, a third fractional amount of the first feature in the given image element, and a fourth fractional amount of the third feature in the given image element; (m) storing the third and fourth fractional amounts in a memory device; (n) repeating steps (k) through (m) for a plurality of image elements in the first image; (o) from first, second, third and fourth fractional amounts for each of a plurality of image elements in the first image, deriving fractional quantities for the first, second, and third features in each a plurality of image elements; and (p) utilizing the fractional quantities derived in step (o) to produce a measurement of a physical parameter of one of the first, second, and third features. - View Dependent Claims (11, 12)
-
Specification