Quantitation and standardization of magnetic resonance measurements

Claim 1

A method of using magnetic resonance imaging (MRI) to produce an image of a body, the method comprising the steps of: using an MRI apparatus to produce a training set comprising one or more training samples, the training set being formed from a plurality of congruent first images of a training region of the body, each first image being produced using an MRI pulse sequence different from the pulse sequences used to produce the other first images, each first image comprising an array of pixels, each training sample comprising a spatially aligned set of pixels from each first image; using an MRI apparatus to produce a test set comprising a plurality of test samples, the test set being formed from a plurality of congruent second images of a test region of the same body, the second images being produced using the same MRI pulse sequences as the first images, each second image comprising an array of pixels, each test sample comprising a spatially aligned set of pixels from each second image; producing similarity data, based on cluster analysis, indicating, for each test sample, the degree of similarity between the test sample and the training samples; and producing a display based upon the similarity data.

Claim 2

A method of using magnetic resonance imaging (MRI) to produce an image of a body, the method comprising the steps of: using an MRI apparatus to produce a training set comprising one or more training samples, the training set being formed from a plurality of congruent first images of a training region of the body, each first image being produced using an MRI pulse sequence different from the pulse sequences used to produce the other first images, each first image comprising an array of pixels, each training sample comprising a spatially aligned set of pixels from each first image; using an MRI apparatus to produce a test set comprising a plurality of test samples, the test set being formed from a plurality of congruent second images of a test region of the same body, the second images being produced using the same MRI pulse sequences as the first images, each second image comprising an array of pixels, each test sample comprising a spatially aligned set of pixels from each second image; producing similarity data, based on cluster analysis, indicating, for each test sample, the degree of similarity between the test sample and the training samples; and producing a display based upon the similarity data.

Claim 4

A method of using magnetic resonance imaging (MRI) to produce an image of a body, the method comprising the steps of: using an MRI apparatus to produce a training set comprising one or more training samples, the training set being formed from a plurality of congruent first images of a training region of the body, each first image being produced using an MRI pulse sequence different from the pulse sequences used to produce the other first images, each first image comprising an array of pixels, each training sample comprising a spatially aligned set of pixels from each first image; using an MRI apparatus to produce a test set comprising a plurality of test samples, the test set being formed from a plurality of congruent second images of a test region of the same body, the second images being produced using the same MRI pulse sequences as the first images, each second image comprising an array of pixels, each test sample comprising a spatially aligned set of pixels from each second image; producing similarity data indicating, for each test sample, the degree of similarity between the test sample and the training samples; and calculating a volume by identifying those pixels having at least a first degree of similarity and multiplying the number of said pixels by the volumes of said body represented by said pixels.

Claim 5

A method of using magnetic resonance imaging (MRI) to produce an image of a body, the method comprising the steps of: using an MRI apparatus to produce a training set comprising one or more training samples, the training set being formed from a plurality of congruent first images of a training region of the body, each first image being produced using an MRI pulse sequence different from the pulse sequences used to produce the other first images, each first image comprising an array of pixels, each training sample comprising a spatially aligned set of pixels from each first image; using an MRI apparatus to produce a test set comprising a plurality of test samples, the test set being formed from a plurality of congruent second images of a test region of the same body, the second images being produced using the same MRI pulse sequences as the first images, each second image comprising an array of pixels, each test sample comprising a spatially aligned set of pixels from each second image; producing similarity data indicating, for each test sample, the degree of similarity between the test sample and the training samples; and calculating a volume by identifying those pixels having at least a first degree of similarity and multiplying the number of said pixels by the volumes of said body represented by said pixels.

Claim 6

A method of using magnetic resonance imaging (MRI) to produce an image of a test object, the method comprising the steps of: using an MRI apparatus to produce a training set comprising one or more training samples, the training set being formed from a plurality of congruent first images of a training region of a first object, each first image being produced using an MRI pulse sequence different from the pulse sequences used to produce the other first images, each first image comprising an array of pixels, each training sample comprising a spatially aligned set of pixels from each first image, at least some of said first images including an image of at least one training set reference object; using an MRI apparatus to produce a test set comprising a plurality of test samples, the test set being formed from a plurality of congruent second images of a test region of the test object, the second images being produced using the same MRI pulse sequences as the first images, each second image comprising an array of pixels, each test sample comprising a spatially aligned set of pixels from each second image, at least some of said second images including an image of at least one test set reference object substantially similar to said test set reference object; correcting at least part of said test set based on differences between said image of said training set reference object and said image of said test set reference object; producing similarity data indicating, for each test sample, the degree of similarity between the test sample and the training samples; and producing a display based upon the similarity data.

Claim 7

A method of using magnetic resonance imaging (MRI) to produce an image of a test object, the method comprising the steps of: using an MRI apparatus to produce a training set comprising one or more training samples, the training set being formed from a plurality of congruent first images of a training region of a first object, each first image being produced using an MRI pulse sequence different from the pulse sequences used to produce the other first images, each first image comprising an array of pixels, each training sample comprising a spatially aligned set of pixels from each first image, at least some of said first images including an image of at least one training set reference object; using an MRI apparatus to produce a test set comprising a plurality of test samples, the test set being formed from a plurality of congruent second images of a test region of the test object, the second images being produced using the same MRI pulse sequences as the first images, each second image comprising an array of pixels, each test sample comprising a spatially aligned set of pixels from each second image, at least some of said second images including an image of at least one test set reference object substantially similar to said test set reference object; correcting at least part of said test set based on differences between said image of said training set reference object and said image of said test set reference object; producing similarity data indicating, for each test sample, the degree of similarity between the test sample and the training samples; and producing a display based upon the similarity data.

Claim 8

A method of using magnetic resonance imaging (MRI) to produce an image of a test object, the method comprising the steps of: using an MRI apparatus to produce a training set comprising one or more training samples, the training set being formed from a plurality of congruent first images of a training region of a first object, each first image being produced using an MRI pulse sequence different from the pulse sequences used to produce the other first images, each first image comprising an array of pixels, each training sample comprising a spatially aligned set of pixels from each first image, at least some of said first images including an image of at least one training set reference object; using an MRI apparatus to produce a test set comprising a plurality of test samples, the test set being formed from a plurality of congruent second images of a test region of the test object, the second images being produced using the same MRI pulse sequences as the first images, each second image comprising an array of pixels, each test sample comprising a spatially aligned set of pixels from each second image, at least some of said second images including an image of at least one test set reference object substantially similar to said test set reference object; producing similarity data indicating, for each test sample, the degree of similarity between the test sample and the training samples; correcting at least part of said similarity data based on differences between said image of said training set reference object and said image of said test set reference object; and producing a display based upon the similarity data.

Claim 9

A method of using magnetic resonance imaging (MRI) to produce an image of a test object, the method comprising the steps of: using an MRI apparatus to produce a training set comprising one or more training samples, the training set being formed from a plurality of congruent first images of a training region of a first object, each first image being produced using an MRI pulse sequence different from the pulse sequences used to produce the other first images, each first image comprising an array of pixels, each training sample comprising a spatially aligned set of pixels from each first image, at least some of said first images including an image of at least one training set reference object; using an MRI apparatus to produce a test set comprising a plurality of test samples, the test set being formed from a plurality of congruent second images of a test region of the test object, the second images being produced using the same MRI pulse sequences as the first images, each second image comprising an array of pixels, each test sample comprising a spatially aligned set of pixels from each second image, at least some of said second images including an image of at least one test set reference object substantially similar to said test set reference object; producing similarity data indicating, for each test sample, the degree of similarity between the test sample and the training samples; correcting at least part of said similarity data based on differences between said image of said training set reference object and said image of said test set reference object; and producing a display based upon the similarity data.

Claim 10

A method of using magnetic resonance imaging (MRI) to produce an image of a test object, the method comprising the steps of: using an MRI apparatus to produce a training set comprising one or more training samples, the training set being formed from a plurality of congruent first images of a training region of a first object, each first image being produced using an MRI pulse sequence different from the pulse sequences used to produce the other first images, each first image comprising an array of pixels, wherein each pixel comprises a pixel value corresponding to the intensity of a magnetic resonance signal from a corresponding position within the object, each training sample comprising a spatially aligned set of pixels from each first image and wherein the training set includes at least one spatial correlation image corresponding to and congruent with one of the first images, the spatial correlation image comprising an array of spatial correlation pixels, each spatial correlation pixel having a pixel value that is determined on the basis of a textural feature, wherein each training sample comprises a spatially aligned set of pixels from each first image and from each first spatial correlation image; using an MRI apparatus to produce a test set comprising a plurality of test samples, the test set being formed from a plurality of congruent second images of a test region of the test object, the second images being produced using the same MRI pulse sequences as the first images, each second image comprising an array of pixels, wherein each pixel comprises a pixel value corresponding to the intensity of a magnetic resonance signal from a corresponding position within the object, each test sample comprising a spatially aligned set of pixels from each second image wherein the test set includes at least one second spatial correlation image corresponding to and congruent with one of the second images, the second spatial correlation image comprising an array of second spatial correlation pixels, each second spatial correlation pixel having a pixel value that is determined on the basis of a textural feature, each test sample comprising a spatially aligned set of pixels from each second image and from each second spatial correlation image; producing similarity data indicating, for each test sample, the degree of similarity between the test sample and the training samples; and producing a display based upon the similarity data.

Claim 11

A method of using magnetic resonance imaging (MRI) to produce an image of a test object, the method comprising the steps of: using an MRI apparatus to produce a training set comprising one or more training samples, the training set being formed from a plurality of congruent first images of a training region of a first object, each first image being produced using an MRI pulse sequence different from the pulse sequences used to produce the other first images, each first image comprising an array of pixels, wherein each pixel comprises a pixel value corresponding to the intensity of a magnetic resonance signal from a corresponding position within the object, each training sample comprising a spatially aligned set of pixels from each first image and wherein the training set includes at least one spatial correlation image corresponding to and congruent with one of the first images, the spatial correlation image comprising an array of spatial correlation pixels, each spatial correlation pixel having a pixel value that is determined on the basis of a textural feature, wherein each training sample comprises a spatially aligned set of pixels from each first image and from each first spatial correlation image; using an MRI apparatus to produce a test set comprising a plurality of test samples, the test set being formed from a plurality of congruent second images of a test region of the test object, the second images being produced using the same MRI pulse sequences as the first images, each second image comprising an array of pixels, wherein each pixel comprises a pixel value corresponding to the intensity of a magnetic resonance signal from a corresponding position within the object, each test sample comprising a spatially aligned set of pixels from each second image wherein the test set includes at least one second spatial correlation image corresponding to and congruent with one of the second images, the second spatial correlation image comprising an array of second spatial correlation pixels, each second spatial correlation pixel having a pixel value that is determined on the basis of a textural feature, each test sample comprising a spatially aligned set of pixels from each second image and from each second spatial correlation image; producing similarity data indicating, for each test sample, the degree of similarity between the test sample and the training samples; and producing a display based upon the similarity data.

Claim 12

A method of using magnetic resonance imaging (MRI) to produce an image of a body, the method comprising the steps of: using an MRI apparatus to produce a training set comprising one or more training samples, the training set being formed from a plurality of congruent first images of a training region of the body, each first image being produced using an MRI pulse sequence different from the pulse sequences used to produce the other first images, each first image comprising an array of pixels, each training sample comprising a spatially aligned set of pixels from each first image; using an MRI apparatus to produce a test set comprising a plurality of test samples, the test set being formed from a plurality of congruent second images of a test region of the same body, the second images being produced using the same MRI pulse sequences as the first images, each second image comprising an array of pixels, each test sample comprising a spatially aligned set of pixels from each second image; producing similarity data indicating, for each test sample, the degree of similarity between the test sample and the training samples; producing a first image based on at least some of said second congruent images producing a second image which displays only those portions of the first image which are within a user-defined similarity threshold of a portion of said training set; and subtracting said second image from said first image to produce a third image.

Claim 13

A method of using magnetic resonance imaging (MRI) to produce an image of a body, the method comprising the steps of: using an MRI apparatus to produce a training set comprising one or more training samples, the training set being formed from a plurality of congruent first images of a training region of the body, each first image being produced using an MRI pulse sequence different from the pulse sequences used to produce the other first images, each first image comprising an array of pixels, each training sample comprising a spatially aligned set of pixels from each first image; using an MRI apparatus to produce a test set comprising a plurality of test samples, the test set being formed from a plurality of congruent second images of a test region of the same body, the second images being produced using the same MRI pulse sequences as the first images, each second image comprising an array of pixels, each test sample comprising a spatially aligned set of pixels from each second image; producing similarity data indicating, for each test sample, the degree of similarity between the test sample and the training samples; producing a first image based on at least some of said second congruent images producing a second image which displays only those portions of the first image which are within a user-defined similarity threshold of a portion of said training set; and subtracting said second image from said first image to produce a third image.