Back to results
US 7,002,345 B2Granted
Synthetic images for a magnetic resonance imaging scanner using linear combinations of source images
AssigneeBoston Medical Center Corporation
InventorHernan Jara
Life of this patent
Estimating the term…
Abstract
The invention consists of three image-postprocessing phases for the purposes of generating high-quality quantitative MR images (proton density (PD), T1, and T2) as well as high-quality virtual MR images with continuously adjustable computer-synthesized contrast weightings, from source images acquired directly with an MRI scanner. Each of the image-postprocessing phases uses one or several new computer algorithms that improve image quality with respect to prior art, including linear-combination-of source-images (LCSI) algorithms for generating PD images and model-conforming algorithms for generating Q-MR images of tissue properties that influence NMR relaxation.
Claims26
A magnetic resonance imaging scanner, comprising:
- a longitudinal magnetic field coil applying a longitudinal magnetic field to a patient;
- at least one gradient magnetic field coil applying a magnetic field gradient to said patient;
- a radio frequency transmitter applying radio frequency electromagnetic energy to said patient;
- a radio frequency receiver coil connected to a radio frequency receiver receiving radio frequency electromagnetic emissions from tissue of said patient;
- control circuits applying a desired pulse sequence of electromagnetic pulses and magnetic field gradients to said patient in order to generate data from which an image of said patient may be constructed;
- a memory storing said data and storing a computer program;
- a CPU executing said computer program, said computer program analyzing said data by performing the following steps,
- a. acquiring a plurality of images by said magnetic resonance imaging scanner using a predefined pulse sequence;
- b. forming a linear combination of said plurality of images, said linear combination formed by summing a product of a coefficient and an image of said plurality of images in order to form a sum of coefficients and images;
- c. evaluating said coefficients in order to achieve a desired effect on said image in order to give said coefficients a specific numerical value;
- and,
- d. synthesizing a desired image using said linear combination of said plurality of images and said specific numerical value of each coefficient.
9 more elements
A magnetic resonance imaging scanner, comprising:
- a longitudinal magnetic field coil applying a longitudinal magnetic field to a patient;
- at least one gradient magnetic field coil applying a magnetic field gradient to said patient;
- a radio frequency transmitter applying radio frequency electromagnetic energy to said patient;
- a radio frequency receiver coil connected to a radio frequency receiver receiving radio frequency electromagnetic emissions from tissue of said patient;
- control circuits applying a desired pulse sequence of electromagnetic pulses and magnetic field gradients to said patient in order to generate data from which an image of said patient may be constructed;
- a memory storing said data and storing a computer program;
- a CPU executing said computer program, said computer program analyzing said data by performing the following steps,
- a. acquiring a plurality of images by said magnetic resonance imaging scanner using a predefined pulse sequence;
- b. forming a linear combination of said plurality of images, said linear combination formed by summing a product of a coefficient and an image of said plurality of images in order to form a sum of coefficients and images;
- c. evaluating said coefficients in order to achieve a desired effect on said image by giving said coefficients a specific numerical value;
- and,
- d. synthesizing a desired image using said linear combination of said plurality of images and said specific numerical value of each coefficient,
- wherein said computer program,
- e. computes tissue values of model parameters on a pixel-by-pixel basis;
- f. sets the tissue value to a predetermined value in response to a measured value of the image giving a calculated value within a specified range of noise value of the image;
- and
- g. sets a value of T1 to zero in the event that an image value in a pixel is less than a noise value.
14 more elements
A magnetic resonance imaging scanner, comprising:
- a longitudinal magnetic field coil applying a longitudinal magnetic field to a patient;
- at least one gradient magnetic field coil applying a magnetic field gradient to said patient;
- a radio frequency transmitter applying radio frequency electromagnetic energy to said patient;
- a radio frequency receiver coil connected to a radio frequency receiver receiving radio frequency electromagnetic emissions from tissue of said patient;
- control circuits applying a desired pulse sequence of electromagnetic pulses and magnetic field gradients to said patient in order to generate data from which an image of said patient may be constructed;
- a memory storing said data and storing a computer program;
- a CPU executing said computer program, said computer program analyzing said data by performing the following steps,
- a. acquiring a plurality of images by said magnetic resonance imaging scanner using a predefined pulse sequence;
- b. forming a linear combination of said plurality of images, said linear combination formed by summing a product of a coefficient and an image of said plurality of images in order to form a sum of coefficients and images;
- c. evaluating said coefficient in order to achieve a desired effect on said image by giving said coefficients a specific numerical value;
- and,
- d. synthesizing a desired image using said linear combination of said plurality of images and said specific numerical value of each coefficient,
- wherein said computer program,
- e. computes tissue values of model parameters on a pixel-by-pixel basis;
- f. sets the tissue value to a predetermined value in response to a measured value of the image giving a calculated value within a specified range of noise value of the image;
- and
- g. sets a value of T2 equal to zero in the event that a pixel value in an image is less than a noise value for that pixel.
14 more elements
A magnetic resonance imaging scanner, comprising:
- a longitudinal magnetic field coil applying a longitudinal magnetic field to a patient;
- at least one gradient magnetic field coil applying a magnetic field gradient to said patient;
- a radio frequency transmitter applying radio frequency electromagnetic energy to said patient;
- a radio frequency receiver coil connected to a radio frequency receiver receiving radio frequency electromagnetic emissions from tissue of said patient;
- control circuits applying a desired pulse sequence of electromagnetic pulses and magnetic field gradients to said patient in order to generate data from which an image of said patient may be constructed;
- a memory storing said data and storing a computer program;
- a CPU executing said computer program, said computer program analyzing said data by performing the following steps,
- a. acquiring a plurality of images by said magnetic resonance imaging scanner using a predefined pulse sequence;
- b. forming a linear combination of said plurality of images, said linear combination formed by summing a product of a coefficient and an image of said plurality of images in order to form a sum of coefficients and images;
- c. evaluating said coefficients in order to achieve a desired effect on said image by giving to give said coefficients a specific numerical value;
- and,
- d. synthesizing a desired image using said linear combination of said plurality of images and said specific numerical value of each coefficient,
- wherein said computer program,
- e. computes tissue values of model parameters on a pixel-by-pixel basis;
- f. sets the tissue value to a predetermined value in response to a measured value of the image giving a calculated value within a specified range of noise value of the image;
- and
- g. sets a value of T2 equal to the value of T2 for water for a pixel, in the event that for said pixel the computed image value:
- IR1_EI−IR2_E1
- is less than a noise value for said pixel.
16 more elements
A magnetic resonance imaging scanner, comprising:
- a longitudinal magnetic field coil applying a longitudinal magnetic field to a patient;
- at least one gradient magnetic field coil applying a magnetic field gradient to said patient;
- a radio frequency transmitter applying radio frequency electromagnetic energy to said patient;
- a radio frequency receiver coil connected to a radio frequency receiver receiving radio frequency electromagnetic emissions from tissue of said patient;
- control circuits applying a desired pulse sequence of electromagnetic pulses and magnetic field gradients to said patient in order to generate data from which an image of said patient may be constructed;
- a memory storing said data and storing a computer program;
- a CPU executing said computer program, said computer program analyzing said data by performing the following steps,
- a. acquiring a plurality of images by said magnetic resonance imaging scanner using a predefined pulse sequence;
- b. forming a linear combination of said plurality of images, said linear combination formed by summing a product of a coefficient and an image of said plurality of images in order to form a sum of coefficients and images;
- c. evaluating said coefficients in order to achieve a desired effect on said image by giving said coefficients a specific numerical value;
- d. synthesizing a desired image using said linear combination of said plurality of images and said specific numerical value of each coefficient;
- e. said computer program,
- e.1 computing on a pixel-by-pixel basis the values of quantitative tissue parameters using an MR physics model;
- e.2 computing noise levels from said plurality of images in order to obtain measured noise levels;
- e.3 executing a conditional statement, said conditional statement comparing a parameter computed from said quantitative tissue parameters with said measured noise levels;
- e.4 setting, in response to said executed conditional statement generating a first result, an image value in a particular pixel equal to a value computed from said MR physics model;
- e.5 setting, in response to said executed conditional statement generating a second result, an image value in a particular pixel equal to a predetermined value.
14 more elements
Dependencies are inferred from the claim text.