# US 12,217,417 B2 · Learning-based domain transformation for medical images

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

- Number: US 12,217,417 B2
- Document ID: US12217417B2
- Status: Granted
- Kind: B2
- Application number: 17/470,076
- Filed: 2021-09-09
- Published: 2023-03-09
- Granted: 2025-02-04
- Earliest claimed: 2021-09-09
- Assignees: GE Precision Healthcare LLC; University of Zurich
- Applicants: GE Precision Healthcare LLC; University of Zurich
- Inventors: Sidharth Abrol; Bipul Das; Vanika Singhal; Amy Deubig; Sandeep Dutta; Daphné Gerbaud; Bianca Sintini; Ronny Büchel; Philipp Kaufmann
- Web page: https://patentgrep.com/patent/US12217417B2

## CPC

- G06T 7/0012: {Biomedical image inspection}
- G06T 12/30: Image post-processing, e.g. metal artefact correction
- G06N 20/00: Machine learning
- G06T 5/70: Denoising; Smoothing
- G06T 7/11: Region-based segmentation
- G06V 10/25: Determination of region of interest [ROI] or a volume of interest [VOI]
- G16H 30/40: for processing medical images, e.g. editing
- G06T 2207/10081: Computed x-ray tomography [CT]
- G06T 2207/20081: Training; Learning
- G06T 2211/441: AI-based methods, deep learning or artificial neural networks
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G16H: HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA

## Abstract

Systems/techniques that facilitate learning-based domain transformation for medical images are provided. In various embodiments, a system can access a medical image. In various aspects, the medical image can depict an anatomical structure according to a first medical scanning domain. In various instances, the system can generate, via execution of a machine learning model, a predicted image based on the medical image. In various aspects, the predicted image can depict the anatomical structure according to a second medical scanning domain that is different from the first medical scanning domain. In some cases, the first and second medical scanning domains can be first and second energy levels of a computed tomography (CT) scanning modality. In other cases, the first and second medical scanning domains can be first and second contrast phases of the CT scanning modality.

## Claims

### Claim 1

A system, comprising:
a processor that executes computer-executable components stored in a computer-readable memory, the computer-executable components comprising:
a receiver component that accesses a medical image, wherein the medical image depicts an anatomical structure according to a first medical scanning domain;
a segmentation component that identifies, via execution of a segmentation model, a region-of-interest in the medical image; and
a transformation component that generates, via execution of a machine learning model, a predicted image based on the medical image, wherein the predicted image depicts the anatomical structure according to a second medical scanning domain that is different from the first medical scanning domain, and wherein the transformation component executes the machine learning model on the region-of-interest and does not execute the machine learning model on a remainder of the medical image.

### Claim 2 (depends on 1)

The system of claim 1, wherein the first medical scanning domain is a first energy level of a computed tomography (CT) scanning modality, and wherein the second medical scanning domain is a second energy level of the CT scanning modality that is different from the first energy level.

### Claim 3 (depends on 1)

The system of claim 1, wherein the first medical scanning domain is a first contrast phase of a computed tomography (CT) scanning modality, and wherein the second medical scanning domain is a second contrast phase of the CT scanning modality that is different from the first contrast phase.

### Claim 4 (depends on 1)

The system of claim 1, wherein the computer-executable components further comprise:
a denoise component that applies one or more denoising techniques to the medical image prior to execution of the machine learning model.

### Claim 5 (depends on 1)

The system of claim 1, wherein the machine learning model is a deep learning regression model.

### Claim 6 (depends on 1)

The system of claim 1, wherein the computer-executable components further comprise:
a training component that trains the machine learning model on a training medical image, wherein the training medical image depicts one or more anatomical structures according to the first medical scanning domain, and wherein the training medical image corresponds to a target annotation image that depicts the one or more anatomical structures according to the second medical scanning domain.

### Claim 7 (depends on 6)

The system of claim 6, wherein the training component registers the training medical image with the target annotation image prior to training the machine learning model.

### Claim 8

A computer-implemented method, comprising:
accessing, by a device operatively coupled to a processor, a medical image, wherein the medical image depicts an anatomical structure according to a first medical scanning domain;
identifying, by the device and via execution of a segmentation model, a region-of-interest in the medical image; and
generating, by the device and via execution of a machine learning model, a predicted image based on the medical image, wherein the predicted image depicts the anatomical structure according to a second medical scanning domain that is different from the first medical scanning domain, and wherein the machine learning model is executed on the region-of-interest and not on a remainder of the medical image.

### Claim 9 (depends on 8)

The computer-implemented method of claim 8, wherein the first medical scanning domain is a first energy level of a computed tomography (CT) scanning modality, and wherein the second medical scanning domain is a second energy level of the CT scanning modality that is different from the first energy level.

### Claim 10 (depends on 8)

The computer-implemented method of claim 8, wherein the first medical scanning domain is a first contrast phase of a computed tomography (CT) scanning modality, and wherein the second medical scanning domain is a second contrast phase of the CT scanning modality that is different from the first contrast phase.

### Claim 11 (depends on 8)

The computer-implemented method of claim 8, further comprising:
applying, by the device, one or more denoising techniques to the medical image prior to execution of the machine learning model.

### Claim 12 (depends on 8)

The computer-implemented method of claim 8, wherein the machine learning model is a deep learning regression model.

### Claim 13 (depends on 8)

The computer-implemented method of claim 8, further comprising:
training, by the device, the machine learning model on a training medical image, wherein the training medical image depicts one or more anatomical structures according to the first medical scanning domain, and wherein the training medical image corresponds to a target annotation image that depicts the one or more anatomical structures according to the second medical scanning domain.

### Claim 14 (depends on 13)

The computer-implemented method of claim 13, further comprising:
registering, by the device, the training medical image with the target annotation image prior to the training.

### Claim 15

A computer program product for facilitating learning-based domain transformation for medical images, the computer program product comprising a computer-readable memory having program instructions embodied therewith, the program instructions executable by a processor to cause the processor to:
access a medical image, wherein the medical image depicts an anatomical structure according to a first medical scanning domain;
identify, via execution of a segmentation model, a region-of-interest in the medical image; and
generate, via execution of a machine learning model, a predicted image based on the medical image, wherein the predicted image depicts the anatomical structure according to a second medical scanning domain that is different from the first medical scanning domain, and wherein the machine learning model is executed on the region-of-interest and not on a remainder of the medical image.

### Claim 16 (depends on 15)

The computer program product of claim 15, wherein the first medical scanning domain is a first energy level of a computed tomography (CT) scanning modality, and wherein the second medical scanning domain is a second energy level of the CT scanning modality that is different from the first energy level.

### Claim 17 (depends on 15)

The computer program product of claim 15, wherein the first medical scanning domain is a first contrast phase of a computed tomography (CT) scanning modality, and wherein the second medical scanning domain is a second contrast phase of the CT scanning modality that is different from the first contrast phase.