# US 2018/0025512 A1 · System and method for segmenting medical image

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

- Number: US 2018/0025512 A1
- Document ID: US20180025512A1
- Status: Published application
- Kind: A1
- Application number: 15/215,575
- Filed: 2016-07-20
- Published: 2018-01-25
- Earliest claimed: 2016-07-20
- Assignees: Shanghai United Imaging Healthcare Co., Ltd.
- Applicants: Shanghai United Imaging Healthcare Co., Ltd.
- Inventors: Wentao ZHU; Tao FENG; Hongdi LI
- Web page: https://patentgrep.com/patent/US20180025512A1

## CPC

- G06T 12/30: Image post-processing, e.g. metal artefact correction
- G01R 33/5608: {Data processing and visualization specially adapted for MR, e.g. for feature analysis and pattern recognition on the basis of measured MR data, segmentation of measured MR data, edge contour detection on the basis of measured MR data, for enhancing measured MR data in terms of signal-to-noise ratio by means of noise filtering or apodization, for enhancing measured MR data in terms of resolution by means for deblurring, windowing, zero filling, or generation of gray-scaled images, colour-coded images or images displaying vectors instead of pixels (image data processing or generation, in general G06T)}
- G06T 7/11: Region-based segmentation
- G06T 12/20: Inverse problem, i.e. transformations from projection space into object space
- G01R 33/481: {MR combined with positron emission tomography [PET] or single photon emission computed tomography [SPECT]}
- G06T 7/187: involving region growing; involving region merging; involving connected component labelling
- G06T 2207/10088: Magnetic resonance imaging [MRI]
- G06T 2207/10104: Positron emission tomography [PET]
- G06T 2207/10108: Single photon emission computed tomography [SPECT]
- G06T 2207/20056: Discrete and fast Fourier transform, [DFT, FFT]
- G06T 2207/30008: Bone
- G06T 2211/424: Iterative
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G01R: MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES (indicating correct tuning of resonant circuits H03J3/12)

## Abstract

A method for segmenting a medical image is disclosed. The method includes acquiring MR image and PET data during a scan of the object, acquiring an air/bone ambiguous region in the MR image, the air/bone ambiguous region including air voxels and bone voxels undistinguished from each other. The method also includes assigning attenuation coefficients to the voxels of the plurality of regions and generating an attenuation map. The method further includes iteratively reconstructing the PET data and the attenuation map to generate a PET image and an estimated attenuation map. The method further includes reassigning attenuation coefficients to the voxels of the air/bone ambiguous region based on the estimated attenuation map, and distinguishing the bone voxels and air voxels in the air/bone ambiguous region.

## Claims

### Claim 1

A method for segmenting a medical image, comprising:
acquiring MR data during a scan of an object using an MR scanner;
reconstructing an MR image using the MR data, wherein the MR image comprises a plurality of voxels;
acquiring PET data during a scan of the object using a PET scanner;
segmenting the MR image into a plurality of regions which includes at least an air/bone ambiguous region, wherein the air/bone ambiguous region including air voxels and bone voxel undistinguished from each other;
assigning attenuation coefficients to the voxels of the plurality of regions and generating an attenuation map, wherein the voxels of the air/bone ambiguous region are assigned specific attenuation coefficients;
iteratively reconstructing the PET data and the attenuation map to generate a PET image and an estimated attenuation map;
reassigning attenuation coefficients to the voxels of the air/bone ambiguous region based on the estimated attenuation map; and
distinguishing the bone voxels and the air voxels in the air/bone ambiguous region.

### Claim 2 (depends on 1)

The method of claim 1, wherein the plurality of regions of the MR image comprises a soft-tissue region, the soft-tissue region comprises voxels of a soft-tissue; and
the air/bone ambiguous region is acquired by excluding the voxels of the soft-tissue regions from the MR image.

### Claim 3 (depends on 2)

The method of claim 2, wherein values of the voxels in the soft-tissue region are similar to each other in the MR image, and values of the voxels in the air/bone ambiguous region are similar to each other in the MR image.

### Claim 4 (depends on 1)

The method of claim 1, wherein the attenuation coefficient of each voxel in the air/bone ambiguous region is updated after each iteration.

### Claim 5 (depends on 1)

The method of claim 1, wherein the specific attenuation coefficients of the voxels in the air/bone ambiguous region are equal to the attenuation coefficients of water.

### Claim 6 (depends on 1)

The method of claim 1, further comprising:
segmenting the air/bone ambiguous region into a first region and a second region, the first region comprises only the bone voxels and the second region comprises only the air voxels.

### Claim 7

A method for segmenting a medical image, including:
acquiring a first type of medical image via reconstructing a first type of data, wherein the first type of medical image comprises a plurality of voxels;
acquiring a second type of data for reconstructing a second type of medical image;
executing preliminary segmentation to the first type of medical image, and obtaining at least an unambiguous region and at least an ambiguous region;
generating an attenuation map by assigning different attenuation coefficients corresponding to voxels belonging to different regions of the first type of image being executed preliminary segmentation;
iteratively estimating the second type of data and the attenuation map to generate the second type of medical image and an estimated attenuation map;
reassigning attenuation coefficients to the voxels of the first type of medical image based on the estimated attenuation map; and
segmenting the first type of medical image into a plurality of sub-regions distinguished from each other.

### Claim 8 (depends on 7)

The method of claim 7, wherein the first type of medical image is an MR image, and the MR image is segmented based on at least one approach selected from a thresholding approach, a region growing approach, a histogram approach, an atlas-guided approach, or a clustering approach.

### Claim 9 (depends on 7)

The method of claim 7, wherein the ambiguous region comprises at least an air/bone ambiguous region, and the unambiguous region comprises at least a soft-tissue region.

### Claim 10 (depends on 9)

The method of claim 9, wherein values of the voxels in the soft-tissue region are similar to each other, and values of the voxels in the air/bone ambiguous region are similar to each other.

### Claim 11 (depends on 10)

The method of claim 10, wherein the voxels of the air/bone ambiguous region are assigned specific attenuation coefficients.

### Claim 12 (depends on 11)

The method of claim 11, wherein the specific attenuation coefficients are equal to attenuation coefficient of water.

### Claim 13 (depends on 7)

The method of claim 7, wherein the first type of medical image is produced by MR device and the second type of medical image is produced by PET device.

### Claim 14 (depends on 9)

The method of claim 9, wherein the unambiguous region comprises a first soft-tissue region and a second soft-tissue region, and the voxels of the first soft-tissue region are assigned a first attenuation coefficient, and the voxels of the second soft-tissue region are assigned a second attenuation coefficient.

### Claim 15 (depends on 14)

The method of claim 14, wherein the first attenuation coefficient or the second attenuation coefficient is updated after each iteration.

### Claim 16 (depends on 9)

The method of claim 9, wherein the attenuation coefficient of each voxel in the air/bone ambiguous region is updated after each iteration.

### Claim 17

A system for segmenting a medical image, the system comprising:
an image processor comprising:
a first reconstruction module configured to acquire an MR image via reconstructing the MR data;
a segmentation module configured to segment the MR image into at least an air/bone ambiguous region and a soft-tissue region, wherein the air/bone ambiguous region comprises air voxels and bone voxels undistinguished from each other, and the soft-tissue region comprises only voxels of soft-tissue;
a second reconstruction module configured to
assign initial attenuation coefficient to each voxel of the MR image being segmented, and generate an attenuation map, wherein voxels of the air/bone ambiguous region are assigned attenuation coefficient of water;
iteratively update the PET data and the attenuation map to generate a PET image and an estimated attenuation map; and,
reassign attenuation coefficient to each voxel of the MR image being segmented based on the estimated attenuation map.

### Claim 18 (depends on 17)

The system of claim 17, wherein the image processor is further configured to segment the air/bone ambiguous region into a first region, a second region, and a third region, wherein the first region comprises only voxels of a rib, the second region comprises only voxels of a spine, and the third region comprises only voxels of a lung.

### Claim 19 (depends on 18)

The system of claim 18, wherein the attenuation coefficient corresponding to the bone voxels in the first region are different from the attenuation coefficient corresponding to the air voxels in the second region.

### Claim 20 (depends on 17)

The system of claim 17, further comprising:
an MR scanner configured to acquire MR data during a first scan of an object;
a PET scanner configured to acquire PET data during a second scan of the object; and
a display configured to simultaneously display the PET image and the MR image in an overlaying manner.
