# US 2004/0017880 A1 · Method and system for low dose image simulation for imaging systems

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

- Number: US 2004/0017880 A1
- Document ID: US20040017880A1
- Status: Published application
- Kind: A1
- Application number: 10/064,586
- Filed: 2002-07-29
- Published: 2004-01-29
- Earliest claimed: 2002-07-29
- Assignees: GE Medical Systems Global Technology Company, LLC
- Inventors: Thomas Louis Toth; Jiang Hsieh; Jianying Li; Scott Matt McOlash
- Web page: https://patentgrep.com/patent/US20040017880A1

## CPC

- G06T 12/30: Image post-processing, e.g. metal artefact correction
- A61B 6/00: Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment (instruments measuring radiation intensity for application in the field of nuclear medicine, e.g. in vivo counting, G01T1/161)
- A61B 6/583: {using calibration phantoms}
- Y10S 378/901: Computer tomography program or processor
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- A61B: DIAGNOSIS; SURGERY; IDENTIFICATION (usefulness limited to only animals A61D)
- Y10S: TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS

## Abstract

A method for generating a simulated patient image is disclosed. In an exemplary embodiment, the method includes obtaining image data from an actual patient image and generating simulated noise data. The image data is then combined with the simulated noise data to create the simulated patient image. In one aspect, scan data from the actual patient image is combined with the generated simulated noise data to create pre-image data, and the pre-image data is then reconstructed to create simulated image data. In another aspect, a set of individual noise pattern images for each a plurality of phantom objects is created. At least one of the individual noise pattern images is selected for combination with the actual patient image. The at least one selected individual noise pattern image is then combined with the actual patient image, thereby creating the simulated patient image.

## Claims

### Claim 1

A method for generating a simulated patient image, the method comprising:
obtaining image data from an actual patient image;
generating simulated noise data; and
combining said image data with said simulated noise data to create the simulated patient image.

### Claim 2 (depends on 1)

The method of claim 1, further comprising:
combining scan data from said actual patient image with said generated simulated noise data to create pre-image data; and
reconstructing said pre-image data to create simulated image data.

### Claim 3 (depends on 2)

The method of claim 2, wherein said simulated noise data is generated through a random number generator in accordance with a Poisson distribution.

### Claim 4 (depends on 3)

The method of claim 3, wherein individual scan data samples from said scan data are each combined with a random noise value generated from said Poisson distribution random number generator, said random noise value first being multiplied by a weighting factor to produce a weighted random noise value.

### Claim 5 (depends on 4)

The method of claim 4, wherein said weighting factor is determined in accordance with the equation:
3 a = β ⁢ D ⁡ ( 1 α - 1 )
wherein a is said weighting factor, β is a scale factor whose value depends on a data acquisition system (DAS) gain and the image processing characteristics, α is a tube current reduction factor relative to a tube current at which said actual patient image was taken, and D is a DAS signal level for a corresponding individual scan data sample.

### Claim 6 (depends on 5)

The method of claim 5, wherein, in addition to said weighting factor, each of said random noise values are further multiplied by an electronic noise scale factor prior to being combined with individual scan data samples, said electronic noise scale factor being determined in accordance with the equation:
σa=aNnP
wherein Nn is said electronic noise scale factor due to non-quantum noise, a is said weighting factor, P is said random noise value generated from said Poisson distribution random number generator, and σa is a standard deviation of said generated simulated noise data to be combined with said actual patient image.

### Claim 7 (depends on 1)

The method of claim 1, further comprising:
creating a set of individual noise pattern images for each a plurality of phantom objects;
selecting at least one of said individual noise pattern images to be combined with said actual patient image; and
combining said at least one selected individual noise pattern image with said actual patient image, thereby creating the simulated patient image.

### Claim 8 (depends on 7)

The method of claim 7, wherein said selecting at least one of said individual noise pattern images is based upon a patient shape and an imaging technique.

### Claim 9 (depends on 7)

The method of claim 7, wherein said at least one of said individual noise pattern images is randomly selected.

### Claim 10 (depends on 9)

The method of claim 9, wherein if more than one of said individual noise pattern images is selected, then said noise pattern images are added together to produce a resultant noise pattern.

### Claim 11 (depends on 10)

The method of claim 10, wherein said combined noise pattern is scaled by a scaling factor, s, in accordance with the equation:
4 s = σ a σ p ; with σ a = σ f 2 - σ 0 2 = σ 0 ⁢ ( 1 α - 1 )
wherein, σa is a standard deviation of said generated simulated noise data to be combined with said actual patient image, ρp is a standard deviation of randomly selected interpolated and summed noise pattern images, σf is a desired standard deviation desired for the simulated patient image, σo is a standard deviation of said actual patient image and α is a tube current reduction factor relative to a tube current at which said actual patient image was take n.

### Claim 12 (depends on 11)

The method of claim 11, wherein said noise pattern images are scaled by the inverse square root of the number of said noise pattern images selected.

### Claim 13

A method for generating a simulated computer tomography (CT) patient image, the method comprising:
obtaining image data from an actual CT patient image taken at a first radiation dose;
generating simulated noise data; and
combining said image data with said simulated noise data to create the simulated patient image;
wherein the simulated image simulates said actual CT patient image taken at a second, reduced radiation dose with respect to said first radiation dose.

### Claim 14 (depends on 13)

The method of claim 13, further comprising:
combining scan data from said actual patient image with said generated simulated noise data to create pre-image data; and
reconstructing said pre-image data to create simulated image data.

### Claim 15 (depends on 14)

The method of claim 14 wherein said simulated noise data is generated through a random number generator in accordance with a Poisson distribution.

### Claim 16 (depends on 15)

The method of claim 15, wherein individual scan data samples from said scan data are each combined with a random noise value generated from said Poisson distribution random number generator, said random noise value first being multiplied by a weighting factor to produce a weighted random noise value.

### Claim 17 (depends on 16)

The method of claim 16, wherein said weighting factor is determined in accordance with the equation:
5 a = β ⁢ D ⁡ ( 1 α - 1 )
wherein a is said weighting factor, β is a scale factor whose value depends on a data acquisition system (DAS) gain and the image processing characteristics, α is a tube current reduction factor relative to a tube current corresponding to said first radiation dose, and D is a DAS signal level for a corresponding individual scan data sample.

### Claim 18 (depends on 17)

The method of claim 17, wherein, in addition to said weighting factor, each of said random noise values are further multiplied by an electronic noise scale factor prior to being combined with individual scan data samples, said electronic noise scale factor being determined in accordance with the equation:
σa=aNnP
wherein Nn is said electronic noise scale factor due to non-quantum noise, a is said weighting factor, P is said random noise value generated from said Poisson distribution random number generator, and σa is a standard deviation of said generated simulated noise data to be combined with said actual patient image.

### Claim 19 (depends on 13)

The method of claim 13, further comprising:
creating a set of individual noise pattern images for each a plurality of phantom objects;
selecting at least one of said individual noise pattern images to be combined with said actual patient image; and
combining said at least one selected individual noise pattern image with said actual patient image, thereby creating the simulated patient image.

### Claim 20 (depends on 19)

The method of claim 19, wherein said selecting at least one of said individual noise pattern images is based upon a patient shape and an imaging technique.

### Claim 21 (depends on 20)

The method of claim 20, wherein said at least one of said individual noise pattern images is randomly selected.

### Claim 22 (depends on 21)

The method of claim 21, wherein if more than one of said individual noise pattern images is selected, then said noise pattern images are added together to produce a resultant noise pattern.

### Claim 23 (depends on 22)

The method of claim 22, wherein said combined noise pattern is scaled by a scaling factor, s ,sin accordance with the equation:
6 s = σ a σ p ; with σ a = σ f 2 - σ 0 2 = σ 0 ⁢ ( 1 α - 1 )
wherein, σa is a standard deviation of said generated simulated noise data to be combined with said actual patient image, σp is a standard deviation of randomly selected interpolated and summed noise pattern images, σf is a desired standard deviation desired for the simulated patient image, σo is a standard deviation of said actual patient image and α is a tube current reduction factor relative to a tube current corresponding to said first radiation dose.

### Claim 24 (depends on 23)

The method of claim 23, wherein said noise pattern images are scaled by the inverse square root of the number of said noise pattern images selected.

### Claim 25

An imaging system, comprising:
a gantry having an x-ray source and a radiation detector array, wherein said gantry defines a patient cavity and wherein said x-ray source and said radiation detector array are rotatingly associated with said gantry so as to be separated by said patient cavity;
a patient support structure movingly associated with said gantry so as to allow communication with said patient cavity; and
a processing device for obtaining image data from an actual patient image;
means for generating simulated noise data; and
means for combining said image data with said simulated noise data to create a simulated patient image.

### Claim 26 (depends on 25)

The imaging system of claim 25, further comprising:
means for combining scan data from said actual patient image with said generated simulated noise data to create pre-image data; and
means for reconstructing said pre-image data to create simulated image data.

### Claim 27 (depends on 26)

The imaging system of claim 26, wherein said simulated noise data is generated through a random number generator in accordance with a Poisson distribution.

### Claim 28 (depends on 27)

The imaging system of claim 27, wherein individual scan data samples from said scan data are each combined with a random noise value generated from said Poisson distribution random number generator, said random noise value first being multiplied by a weighting factor to produce a weighted random noise value.

### Claim 29 (depends on 28)

The imaging system of claim 28, wherein said weighting factor is determined in accordance with the equation:
7 a = β ⁢ D ⁡ ( 1 α - 1 )
wherein a is said weighting factor, β is a scale factor whose value depends on a data acquisition system (DAS) gain and the image processing characteristics, α is a tube current reduction factor relative to a tube current at which said actual patient image was taken, and D i s a DAS signal level for a corresponding individual scan data sample.

### Claim 30 (depends on 29)

The imaging system of claim 29, wherein, in addition to said weighting factor, each of said random noise values are further multiplied by an electronic noise scale factor prior to being combined with individual scan data samples, said electronic noise scale factor being determined in accordance with the equation:
σa=aNnP
wherein Nn is said electronic noise scale factor due to non-quantum noise, as said weighting factor, P is said random noise value generated from said Poisson distribution random number generator, and σa is a standard deviation of said generated simulated noise data to be combined with said actual patient image.

### Claim 31 (depends on 25)

The imaging system of claim 25, further comprising:
means for creating a set of individual noise pattern images for each a plurality of phantom objects;
means for selecting at least one of said individual noise pattern images to be combined with said actual patient image; and
means for combining said at least one selected individual noise pattern image with said actual patient image, thereby creating the simulated patient image.

### Claim 32 (depends on 31)

The imaging system of claim 31, wherein said means for selecting at least one of said individual noise pattern images is based upon a patient shape and an imaging technique.

### Claim 33 (depends on 31)

The imaging system of claim 31, wherein said at least one of said individual noise pattern images is randomly selected.

### Claim 34 (depends on 33)

The imaging system of claim 33, wherein if more than one of said individual noise pattern images is selected, then said noise pattern images are added together to produce a resultant noise pattern.

### Claim 35 (depends on 34)

The imaging system of claim 34, wherein said combined noise pattern is scaled by a scaling factor, ,sin accordance with the equation:
8 s = σ a σ p ; with σ a = σ f 2 - σ 0 2 = σ 0 ⁢ ( 1 α - 1 )
wherein, σa is a standard deviation of said generated simulated noise data to be combined with said actual patient image, σp is a standard deviation of randomly selected interpolated and summed noise pattern images, σf is a desired standard deviation desired for the simulated patient image, σo is a standard deviation of said actual patient image and α is a tube current reduction factor relative to a tube current at which said actual patient image was taken.

### Claim 36 (depends on 35)

The imaging system of claim 35, wherein said noise pattern images are scaled by the inverse square root of the number of said noise pattern images selected.

### Claim 37

A storage medium, comprising:
a machine readable computer program code for generating a simulated patient image; and
instructions for causing a computer to implement a method, the method further comprising:
obtaining image data from an actual patient image;
generating simulated noise data; and
combining said image data with said simulated noise data to create the simulated patient image.

### Claim 38 (depends on 37)

The storage medium of claim 37, further comprising:
combining scan data from said actual patient image with said generated simulated noise data to create pre-image data; and
reconstructing said pre-image data to create simulated image data.

### Claim 39 (depends on 38)

The storage medium of claim 38, wherein said simulated noise data is generated through a random number generator in accordance with a Poisson distribution.

### Claim 40 (depends on 39)

The storage medium of claim 39, wherein individual scan data samples from said scan data are each combined with a random noise value generated from said Poisson distribution random number generator, said random noise value first being multiplied by a weighting factor to produce a weighted random noise value.

### Claim 41 (depends on 40)

The storage medium of claim 40, wherein said weighting factor is determined in accordance with the equation:
9 a = β ⁢ D ⁡ ( 1 α - 1 )
wherein a is said weighting factor, β is a scale factor whose value depends on a data acquisition system (DAS) gain and the image processing characteristics, α is a tube current reduction factor relative to a tube current at which said actual patient image was taken, and D is a DAS signal level for a corresponding individual scan data sample.

### Claim 42 (depends on 41)

The storage medium of claim 41, wherein, in addition to said weighting factor, each of said random noise values are further multiplied by an electronic noise scale factor prior to being combined with individual scan data samples, said electronic noise scale factor being determined in accordance with the equation:
σa=aNnP
wherein Nn is said electronic noise scale factor due to non-quantum noise, a is said weighting factor, P is said random noise value generated from said Poisson distribution random number generator, and σa is a standard deviation of said generated simulated noise data to be combined with said actual patient image.

### Claim 43 (depends on 37)

The storage medium of claim 37, further comprising:
creating a set of individual noise pattern images for each a plurality of phantom objects;
selecting at least one of said individual noise pattern images to be combined with said actual patient image; and
combining said at least one selected individual noise pattern image with said actual patient image, thereby creating the simulated patient image.

### Claim 44 (depends on 43)

The storage medium of claim 43, wherein said selecting at least one of said individual noise pattern images is based upon a patient shape and an imaging technique.

### Claim 45 (depends on 43)

The storage medium of claim 43, wherein said at least one of said individual noise pattern images is randomly selected.

### Claim 46 (depends on 45)

The storage medium of claim 45, wherein if more than one of said individual noise pattern images is selected, then said noise pattern images are added together to produce a resultant noise pattern.

### Claim 47 (depends on 46)

The storage medium of claim 46, wherein said combined noise pattern is scaled by a scaling factor, ,sin accordance with the equation:
10 s = σ a σ p ; with σ a = σ f 2 - σ 0 2 = σ 0 ⁢ ( 1 α - 1 )
wherein, σa is a standard deviation of said generated simulated noise data to be combined with said actual patient image, σp is a standard deviation of randomly selected interpolated and summed noise pattern images, or σf is a desired standard deviation desired for the simulated patient image, σo is a standard deviation of said actual patient image and α is a tube current reduction factor relative to a tube current at which said actual patient image was taken.

### Claim 48 (depends on 47)

The storage medium of claim 47, wherein said noise pattern images are scaled by the inverse square root of the number of said noise pattern images selected.

### Claim 49

A computer data signal, comprising:
code configured to cause a processor to implement a method for generating a simulated patient image, the method further comprising:
obtaining image data from an actual patient image;
generating simulated noise data; and
combining said image data with said simulated noise data to create the simulated patient image.

### Claim 50 (depends on 37)

The computer data signal of claim 37, further comprising:
combining scan data from said actual patient image with said generated simulated noise data to create pre-image data; and reconstructing said pre-image data to create simulated image data.

### Claim 51 (depends on 50)

The computer data signal of claim 50, wherein said simulated noise data is generated through a random number generator in accordance with a Poisson distribution.

### Claim 52 (depends on 51)

The computer data signal of claim 51, wherein individual scan data samples from said scan data are each combined with a random noise value generated from said Poisson distribution random number generator, said random noise value first being multiplied by a weighting factor to produce a weighted random noise value.

### Claim 53 (depends on 52)

The computer data signal of claim 52, wherein said weighting factor is determined in accordance with the equation:
11 a = β ⁢ D ⁡ ( 1 α - 1 )
wherein a is said weighting factor, β is a scale factor whose value depends on a data acquisition system (DAS) gain and the image processing characteristics, α is a tube current reduction factor relative to a tube current at which said actual patient image was taken, and D is a DAS signal level for a corresponding individual scan data sample.

### Claim 54 (depends on 53)

The computer data signal of claim 53, wherein, in addition to said weighting factor, each of said random noise values are further multiplied by an electronic noise scale factor prior to being combined with individual scan data samples, said electronic noise scale factor being determined in accordance with the equation:
σa=aNnP
wherein Nn is said electronic noise scale factor due to non-quantum noise, as said weighting factor, P is said random noise value generated from said Poisson distribution random number generator, and σa is a standard deviation of said generated simulated noise data to be combined with said actual patient image.

### Claim 55 (depends on 49)

The computer data signal of claim 49, further comprising:
creating a set of individual noise pattern images for each a plurality of phantom objects;
selecting at least one of said individual noise pattern images to be combined with said actual patient image; and
combining said at least one selected individual noise pattern image with said actual patient image, thereby creating the simulated patient image.

### Claim 56 (depends on 55)

The computer data signal of claim 55, wherein said selecting at least one of said individual noise pattern images is based upon a patient shape and an imaging technique.

### Claim 57 (depends on 55)

The computer data signal of claim 55, wherein said at least one of said individual noise pattern images is randomly selected.

### Claim 58 (depends on 57)

The computer data signal of claim 57, wherein if more than one of said individual noise pattern images is selected, then said noise pattern images are added together to produce a resultant noise pattern.

### Claim 59 (depends on 58)

The computer data signal of claim 58, wherein said combined noise pattern is scaled by a scaling factor, s, in accordance with the equation:
12 s = σ a σ p ; with ⁢ σ a = σ f 2 - σ o 2 = σ o ⁢ ( 1 α - 1 )
wherein, σa is a standard deviation of said generated simulated noise data to be combined with said actual patient image, σp is a standard deviation of randomly selected interpolated and summed noise pattern images, σf is a desired standard deviation desired for the simulated patient image, σo is a standard deviation of said actual patient image and α is a tube current reduction factor relative to a tube current at which said actual patient image was taken.

### Claim 60 (depends on 59)

The storage medium of claim 59, wherein said noise pattern images are scaled by the inverse square root of the number of said noise pattern images selected.
