# US 2023/0410259 A1 · Noise preserving models and methods for resolution recovery of x-ray computed tomography images

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

- Number: US 2023/0410259 A1
- Document ID: US20230410259A1
- Status: Published application
- Kind: A1
- Application number: 17/807,779
- Filed: 2022-06-20
- Published: 2023-12-21
- Earliest claimed: 2022-06-20
- Assignees: GE Precision Healthcare LLC; Purdue Research Foundation; University of Notre Dame du Lac
- Applicants: GE Precision Healthcare LLC; Purdue Research Foundation; University of Notre Dame du Lac
- Inventors: Roman Melnyk; Madhuri Mahendra Nagare; Jie Tang; Obaidullah Rahman; Brian E Nett; Ken Sauer; Charles Addison Bouman, Jr.
- Web page: https://patentgrep.com/patent/US20230410259A1

## CPC

- G06T 5/70: Denoising; Smoothing
- A61B 6/032: {Transmission computed tomography [CT]}
- A61B 6/5258: {involving detection or reduction of artifacts or noise}
- G06T 5/60: using machine learning, e.g. neural networks
- G06T 5/73: Deblurring; Sharpening
- G06T 7/0012: {Biomedical image inspection}
- G06T 2207/10081: Computed x-ray tomography [CT]
- G06T 2207/20081: Training; Learning
- G06T 2207/20084: Artificial neural networks [ANN]
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- A61B: DIAGNOSIS; SURGERY; IDENTIFICATION (usefulness limited to only animals A61D)

## Abstract

Noise preserving models and methods for resolution recovery of x-ray computed tomography (e.g., using a computerized tool) are enabled. For example, a system can comprise: a memory that stores computer executable components, and a processor that executes the computer executable components stored in the memory, wherein the computer executable components comprise: a pair generation component that generates a pair of images, the pair of images comprising an input image and a ground truth image, a training component that trains a machine learning based sharpening algorithm by approximately minimizing a loss function that determines an error between a sharpened image and the ground truth image, and a sharpening component that, using the sharpening algorithm, sharpens the input image to generate the sharpened image, wherein the sharpened image comprises a second noise that is similar in intensity to a first noise of the input image.

## Claims

### Claim 1

A system, comprising:
a memory that stores computer executable components; and
a processor that executes the computer executable components stored in the memory, wherein the computer executable components comprise:
a pair generation component that generates a pair of images, the pair of images comprising an input image and a ground truth image;
a training component that trains a machine learning based sharpening algorithm by approximately minimizing a loss function that determines an error between a sharpened image and the ground truth image; and
a sharpening component that, using the sharpening algorithm, sharpens the input image to generate the sharpened image, wherein the sharpened image comprises a second noise that is similar in intensity to a first noise of the input image.

### Claim 2 (depends on 1)

The system of claim 1, wherein an extent of the sharpening of the input image is determined, by the sharpening component and using the sharpening algorithm, based on an adjustable sharpening parameter.

### Claim 3 (depends on 1)

The system of claim 1, wherein the computer executable components further comprise:
a noise addition component that adds noise to the pair of images, wherein the noise is respectively scaled for the sharpened image and the input image so that the second noise is similar in intensity to the first noise.

### Claim 4 (depends on 3)

The system of claim 3, wherein an input image scaling factor for the added noise, applicable to the input image, is based on a first adjustable parameter.

### Claim 5 (depends on 3)

The system of claim 3, wherein a ground truth scaling factor for the added noise, applicable to the ground truth image, is based on a second adjustable parameter.

### Claim 6 (depends on 3)

The system of claim 3, wherein one or more scaling factors for the added noise are determined, by the noise addition component, to maintain the second noise similar in intensity to the first noise.

### Claim 7 (depends on 3)

The system of claim 3, wherein the added noise is determined, by the noise addition component, using a water phantom.

### Claim 8 (depends on 3)

The system of claim 3, wherein the added noise is determined, by the noise addition component, by subtracting averaged, high-resolution noise from individual, high-resolution noise.

### Claim 9 (depends on 1)

The system of claim 1, wherein the computer executable components further comprise:
a post-processing component that reduces artifacts in the sharpened image based on a local edge magnitude of the sharpened image.

### Claim 10 (depends on 1)

The system of claim 1, wherein the computer executable components further comprise:
a post-processing component that reduces artifacts in the sharpened image based on Hounsfield Unit values of the sharpened image.

### Claim 11

A method, comprising:
generating, by a system comprising a processor, a pair of images, the pair of images comprising an input image and a ground truth image;
training, by the system, a machine learning based sharpening algorithm by approximately minimizing a loss function that determines an error between a sharpened image and the ground truth image; and
using the sharpening algorithm, sharpening, by the system, the input image to generate the sharpened image, wherein the sharpened image comprises a second noise that is similar in intensity to a first noise of the input image.

### Claim 12 (depends on 11)

The method of claim 11, wherein an extent of the sharpening of the input image is determined, using the sharpening algorithm, based on an adjustable sharpening parameter.

### Claim 13 (depends on 11)

The method of claim 11, further comprising:
adding, by the system, noise to the pair of images, wherein the noise is respectively scaled for the sharpened image and the input image so that the second noise is similar in intensity to the first noise.

### Claim 14 (depends on 13)

The method of claim 13, wherein an input image scaling factor for the added noise, applicable to the input image, is based on a first adjustable parameter, and wherein a ground truth scaling factor for the added noise, applicable to the ground truth image, is based on a second adjustable parameter.

### Claim 15

A non-transitory machine-readable storage medium, comprising executable instructions that, when executed by a processor, facilitate performance of operations, comprising:
generating a pair of images, the pair of images comprising an input image and a ground truth image;
training a machine learning based sharpening algorithm by approximately minimizing a loss function that determines an error between a sharpened image and the ground truth image;
using the sharpening algorithm, sharpening the input image to generate the sharpened image, wherein the sharpened image comprises a second noise that is similar in intensity to a first noise of the input image; and
adding noise to the pair of images, wherein the noise is respectively scaled for the sharpened image and the input image so that the second noise is similar in intensity to the first noise.

### Claim 16 (depends on 15)

The non-transitory machine-readable storage medium of claim 15, wherein one or more scaling factors for the added noise are determined to maintain the second noise similar in intensity to the first noise.

### Claim 17 (depends on 15)

The non-transitory machine-readable storage medium of claim 15, wherein the added noise is determined using a water phantom.

### Claim 18 (depends on 15)

The non-transitory machine-readable storage medium of claim 15, wherein the added noise is determined by subtracting averaged, high-resolution noise from individual, high-resolution noise.

### Claim 19 (depends on 15)

The non-transitory machine-readable storage medium of claim 15, wherein the operations further comprise:
reducing one or more artifacts in the sharpened image based on a local edge magnitude of the sharpened image.

### Claim 20 (depends on 15)

The non-transitory machine-readable storage medium of claim 15, wherein the operations further comprise:
reducing one or more artifacts in the sharpened image based on Hounsfield Unit values of the sharpened image.
