Method for determining a property map of an object, particularly of a living being, based on at least a first image, particularly a magnetic resonance image
First Claim
1. A method for determining a property map of an object based on at least a first magnetic resonance (MR) image of the object, comprising:
- defining a structure of reference pairs, wherein each reference pair has at least first and second entries, wherein each of said first entries represents an unknown property value, and wherein each of said second entries represents a group of image points belonging together, which are extracted from said first MR image recorded by a scanner, wherein the group of image points includes at least one interesting image point which corresponds to one of the unknown property-values, wherein each unknown property value represents an attenuation value;
providing a plurality of training pairs, wherein a structure of the training pairs corresponds to the structure of said reference pairs, and wherein first and second entries of the respective training pairs are known; and
determining an assignment between the first entries and the second entries of the training pairs by machine learning, and predicting an attenuation value for an arbitrary image point which corresponds to the at least one interesting image point of the first MR image by applying the determined assignment to one of the second entries of the reference pairs including said arbitrary image point.
1 Assignment
0 Petitions
Accused Products
Abstract
It is disclosed a system and method (12) for determining a property map (82) of an object, particularly a human being, based on at least a first image (84), particularly an magnetic resonance (MR) image, of the object. In the method (12), a structure of reference pairs is defined in a first step (96), wherein each reference pair (16-26) comprises at least two entries (62). The first entry represents a property value, particularly an attenuation value. The second entry (62) preferably represents a group of image points (67) belonging together, which is extracted particularly from MR images (28) and comprises an interesting image point corresponding to the property value. In another step (98) of the method (12) a plurality of training pairs (16-26) is provided. A structure of the training pairs (16-26) corresponds to the structure of reference pairs, and the entries of respective training pairs (16-26) are known. In another step (100) of the method (12), an assignment between the first entries and the other entries (62-66) of the training pairs (16-26) is determined by machine learning, thus allowing prediction of a property value (88) corresponding to an arbitrary point (90) of the first image (84).
14 Citations
19 Claims
-
1. A method for determining a property map of an object based on at least a first magnetic resonance (MR) image of the object, comprising:
-
defining a structure of reference pairs, wherein each reference pair has at least first and second entries, wherein each of said first entries represents an unknown property value, and wherein each of said second entries represents a group of image points belonging together, which are extracted from said first MR image recorded by a scanner, wherein the group of image points includes at least one interesting image point which corresponds to one of the unknown property-values, wherein each unknown property value represents an attenuation value; providing a plurality of training pairs, wherein a structure of the training pairs corresponds to the structure of said reference pairs, and wherein first and second entries of the respective training pairs are known; and determining an assignment between the first entries and the second entries of the training pairs by machine learning, and predicting an attenuation value for an arbitrary image point which corresponds to the at least one interesting image point of the first MR image by applying the determined assignment to one of the second entries of the reference pairs including said arbitrary image point. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17)
-
-
18. A system for determining a property map of an object based on at least a first magnetic resonance (MR) image of the object, comprising:
-
an imaging system; a computer system connected to the imaging system; wherein the computer system is configured to execute a series of steps to determine the property map of the object, the steps comprising; defining a structure of reference pairs, wherein each reference pair has at least first and second entries, wherein each of said first entries represents an unknown property value, and wherein each of said second entries represents a group of image points belonging together, which are extracted from said first MR image recorded by a scanner, wherein the group of image points includes at least one interesting image point which corresponds to one of the unknown property values, wherein each unknown property value represents an attenuation value; providing a plurality of training pairs, wherein a structure of the training pairs corresponds to the structure of said reference pairs, and wherein first and second entries of the respective training pairs are known; and determining an assignment between the first entries and the second entries of the training pairs by machine learning, and predicting an attenuation value for an arbitrary image point which corresponds to the at least one interesting image point of the first MR image by applying the determined assignment to one of the second entries of the reference pairs including said arbitrary image point.
-
-
19. A non-transitory computer readable medium comprising a set of instructions that when executed by a computer processor, result in a series of steps being performed to determine a property map of an object based on at least a first magnetic resonance (MR) image of the object, the steps comprising:
-
defining a structure of reference pairs, wherein each reference pair has at least first and second entries, wherein each of said first entries represents an unknown property value, and wherein each of said second entries represents a group of image points belonging together, which are extracted from said first MR image recorded by a scanner, wherein the group of image points includes at least one interesting image point which corresponds to one of the unknown property values, wherein each unknown property value represents an attenuation value; providing a plurality of training pairs, wherein a structure of the training pairs corresponds to the structure of said reference pairs, and wherein first and second entries of the respective training pairs are known; and determining an assignment between the first entries and the second entries of the training pairs by machine learning, and predicting an attenuation value for an arbitrary image point which corresponds to the at least one interesting image point of the first MR image by applying the determined assignment to one of the second entries of the reference pairs including said arbitrary image point.
-
Specification