# US 2015/0146845 A1 · Methods and systems for performing model-based image processing

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

- Number: US 2015/0146845 A1
- Document ID: US20150146845A1
- Status: Published application
- Kind: A1
- Application number: 14/092,383
- Filed: 2013-11-27
- Published: 2015-05-28
- Earliest claimed: 2013-11-27
- Assignees: General Electric Company
- Applicants: General Electric Company
- Inventors: Jean-Baptiste Thibault; Ruoqiao Zhang; Charles Bouman; Ken Sauer
- Web page: https://patentgrep.com/patent/US20150146845A1

## CPC

- G06T 12/20: Inverse problem, i.e. transformations from projection space into object space
- A61B 6/032: {Transmission computed tomography [CT]}
- A61B 6/5205: {involving processing of raw data to produce diagnostic data}
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- A61B: DIAGNOSIS; SURGERY; IDENTIFICATION (usefulness limited to only animals A61D)

## Abstract

Methods and systems for model-based image processing are provided. One method includes selecting at least one reference image from a plurality of reference images, partitioning the at least one reference image into a plurality of patches, generating a probability distribution for each of the patches, and generating a model of a probability distribution for the at least one reference image using the probability distributions for each of the patches.

## Claims

### Claim 1

A method of generating a model for image processing, said method comprising:
selecting at least one reference image from a plurality of reference images;
partitioning the at least one reference image into a plurality of patches;
generating a probability distribution for each of the patches; and
generating a model of a probability distribution for the at least one reference image using the probability distributions for each of the patches.

### Claim 2 (depends on 1)

The method of claim 1, further comprising reconstructing one or more images from raw data using the generated model, wherein the one or more images comprise at least one of a two-dimensional image or a three-dimensional image.

### Claim 3 (depends on 1)

The method of claim 1, further comprising performing image processing on one or more images by applying the generated model to the one or more images.

### Claim 4 (depends on 1)

The method of claim 1, wherein the selecting comprises selecting an image volume defined by at least some of the plurality of reference images.

### Claim 5 (depends on 1)

The method of claim 1, further comprising modeling a multi-material density and a spatial texture of the reference image using the generated model.

### Claim 6 (depends on 5)

The method of claim 5, further comprising modeling the multi-material density concurrently with the spatial texture.

### Claim 7 (depends on 1)

The method of claim 1, wherein the model includes a joint model of a density and a texture of the reference image, the method further comprising:
incorporating the joint model into a model based iterative reconstruction (MBIR) algorithm; and
reconstructing the plurality of images using the incorporated model.

### Claim 8 (depends on 1)

The method of claim 1, wherein generating the model comprises:
calculating a density value for each of the patches; and
merging the density values to generate a model of the reference image, the model including a texture and a density of the reference image.

### Claim 9 (depends on 1)

The method of claim 1, thither comprising using a mixture distribution to model the probability distribution of the patches.

### Claim 10 (depends on 1)

The method of claim 1, further comprising using a Gaussian mixture distribution to model the probability distribution of the patches.

### Claim 11

A non-transitory computer readable medium programmed to instruct a computer of a Computed Tomography (CT) system to:
receive a user input selecting at least one reference image from a plurality of reference images;
automatically partition the at least one reference image into a plurality of patches;
generate a probability distribution for each of the patches; and
generate a model of a probability distribution for the at least one reference image using the probability distributions for each of the patches.

### Claim 12 (depends on 11)

The non-transitory computer readable medium of claim 11, being further programmed to reconstruct one or more images from raw data using the generated model, wherein the one or more images comprise at least one of a two-dimensional image or a three-dimensional image.

### Claim 13 (depends on 11)

The non-transitory computer readable medium of claim 11, being further programmed to perform image processing on one or more images by applying the generated model to the one or more images.

### Claim 14 (depends on 11)

The non-transitory computer readable medium of claim 11, being further programmed to model a multi-material density and a spatial texture of the reference image using the generated model.

### Claim 15 (depends on 11)

The non-transitory computer readable medium of claim 11, being further programmed to model the multi-material density concurrently with the spatial texture.

### Claim 16 (depends on 11)

The non-transitory computer readable medium of claim 11, wherein the model includes a joint model of a density and a texture of the reference image, the computer readable medium being further programmed to:
incorporate the joint model into a model based iterative reconstruction (MBIR) algorithm; and
reconstruct the plurality of images using the incorporated model.

### Claim 17 (depends on 11)

The non-transitory computer readable medium of claim 11, to generate the model, being further programmed to:
calculate a density value for each of the patches; and
merge the density values to generate a model of the reference image, the model including a texture and a density of the reference image.

### Claim 18 (depends on 11)

The non-transitory computer readable medium of claim 11, being further programmed to use a mixture distribution to model the probability distribution of the patches.

### Claim 19 (depends on 11)

The non-transitory computer readable medium of claim 11, being further programmed to use a Gaussian mixture distribution to model the probability distribution of the patches.

### Claim 20

A computed tomography (CT) imaging system comprising:
a detector configured to output x-ray attenuation information; and
a processor coupled to the detector, the processor being configured to:
receive a user input selecting at least one reference image from a plurality of reference images;
automatically partition the at least one reference image into a plurality of patches;
generate a probability distribution for each of the patches; and
generate a model of a probability distribution for the at least one reference image using the probability distributions for each of the patches.

### Claim 21 (depends on 20)

The CT system of claim 20, wherein the processor is further programmed to receive the x-ray attenuation information as raw data and reconstruct one or more images from the raw data using the generated model, wherein the one or more images comprise at least one of a two-dimensional image or a three-dimensional image.

### Claim 22 (depends on 20)

The CT system of claim 20, wherein the processor is further programmed to perform image processing on one or more images reconstructed from the x-ray attenuation information by applying the generated model to the one or more images.

### Claim 23 (depends on 20)

The CT system of claim 20, wherein the processor is further programmed to model a multi-material density and a spatial texture of the reference image using the generated model.

### Claim 24 (depends on 20)

The CT system of claim 20, wherein the processor is further programmed to model the multi-material density concurrently with the spatial texture.

### Claim 25 (depends on 20)

The CT system of claim 20, wherein the model includes a joint model of a density and a texture of the reference image, the processor is further programmed to:
incorporate the joint model into a model based iterative reconstruction (MBIR) algorithm; and
reconstruct the plurality of images using the incorporated model.

### Claim 26 (depends on 20)

The CT system of claim 20, wherein the processor, to generate the model, is further programmed to
calculate a density value for each of the patches; and
merge the density values to generate a model of the reference image, the model including a texture and a density of the reference image.

### Claim 27 (depends on 20)

The CT system of claim 20, wherein the processor is further programmed to use a Gaussian mixture distribution to model the probability distribution of the patches.
