KR-20260066172-A - Medical imaging systems for reducing radiation exposure

KR20260066172AKR 20260066172 AKR20260066172 AKR 20260066172AKR-20260066172-A

Abstract

Methods, systems, and devices are described herein, which are intended to reduce inadvertent exposure to harmful radiation by optimizing medical imaging processes using processes and machine learning techniques. A machine learning model may be trained to output recommended operating parameter settings for a medical imaging device. Available operating parameters of a medical imaging device may be determined, and patient data may be received. The patient data and the available operating parameters may be used as input to a trained machine learning model, which may output recommended operating parameter settings. Subsequently, this output may be used to transmit data to the medical imaging device that causes modification of the operating parameters of the medical imaging device, along with other calculations. Metadata corresponding to one or more images acquired by the medical imaging device may be received, and the trained machine learning model may be further trained based on said metadata.

Inventors

킹마 필립 알.
웨버 레트리샤

Assignees

레덕션, 아이엔씨.

Dates

Publication Date: 20260512
Application Date: 20240828
Priority Date: 20230918

Claims (20)

In a computing device configured to use machine learning techniques to optimize medical image processing to reduce unintended radiation exposure, One or more processors; and The computing device includes memory for storing instructions, and when the instructions are executed by the one or more processors: A machine learning model is generated by training a machine learning model to output recommended medical imaging device operation parameter settings using training data including a history of medical imaging device operation parameter settings, patient data, and image results, and training the machine learning model includes modifying one or more weights of one or more nodes of an artificial neural network based on the training data; To receive the indication of the medical imaging device; Based on the display of the medical imaging device, the available operating parameters of the medical imaging device are determined; To receive patient data; To provide data corresponding to the patient data and available operating parameters of the medical imaging device to one or more input nodes of the above-mentioned trained machine learning model; Receiving one or more recommended action parameter settings through one or more output nodes of the above-mentioned trained machine learning model; Data that causes modification of the operating parameters of the medical imaging device based on one or more recommended operating parameter settings is transmitted to the medical imaging device; Receiving metadata corresponding to one or more images captured by the medical imaging device; A computing device that enables further training of the trained machine learning model based on the above metadata.
In Article 1, When the above instructions are executed by the above one or more processors, the computing device: A computing device that determines one or more additional modifications made to the operation parameters of the medical imaging device, and, when the instructions are executed by the one or more processors, causes the computing device to further train the trained machine learning model based on the one or more additional modifications.
In Article 1, A computing device in which, when the above instructions are executed by the one or more processors, the computing device processes the one or more images to identify image quality for each of the one or more images and determines the metadata, thereby enabling the computing device to receive the metadata corresponding to the one or more images captured by the medical imaging device.
In Article 1, When the above instructions are executed by one or more processors, the computing device: The above patient data; and Based on the above one or more recommended operation parameter settings, A computing device that causes the computing device to transmit data that causes the modification of the operation parameters of the medical imaging device by determining the above data.
In Article 1, When the above instructions are executed by the above one or more processors, the computing device: To display a user interface through a user device; A computing device that enables the computing device to receive patient data by receiving patient weight and patient height through the above user interface.
In Article 1, When the above instructions are executed by the above one or more processors, the computing device: A computing device that enables the computing device to determine available operating parameters of the medical imaging device by querying a database of device identification and corresponding operating parameters based on the display of the medical imaging device.
In Article 1, When the above instructions are executed by the above one or more processors, the computing device: One or more images captured by the medical imaging device; Whether a pulse setting is enabled during the capture of one or more of the above images; or A computing device that allows the computing device to receive the metadata corresponding to one or more images captured by the medical imaging device by determining one or more of whether an automatic setting is enabled during the capture of one or more of the above images.
In Article 1, The above data is a computing device configured to allow the medical imaging device to turn off the automatic function and capture multiple images at a predetermined frequency.
In Article 1, The above instructions, when executed by the one or more processors, cause the computing device to receive patient data by causing the computing device to receive said patient data through the user interface of the user device, said patient data represents the patient's weight and the patient's height, and the above instructions, when executed by the one or more processors, cause the computing device: An equalizing quotient (EQ) is determined, wherein the EQ is calculated by 1) multiplying X by the patient's weight, where X is between 1 and 10, preferably less than 4, and most preferably between 1 and 3, and then dividing the result by the patient's height, or 2) based on the patient's Body Mass Index (BMI); A computing device that transmits data causing modification of the operation parameters of the medical imaging device by using the above EQ to determine the above operation parameter settings.
In Article 1, The above data is a computing device configured to cause the medical imaging device to disable automatic exposure control settings.
In a method of using machine learning techniques to optimize medical image processing to reduce unintended harmful radiation exposure, A step of generating a trained machine learning model by training a machine learning model to output recommended medical imaging device operation parameter settings using training data including a history of medical imaging device operation parameter settings, patient data, and image results, wherein the step of training the machine learning model includes a step of modifying one or more weights of one or more nodes of an artificial neural network based on the training data; Step of receiving a display of a medical imaging device; A step of determining available operating parameters of the medical imaging device based on the display of the medical imaging device; Step of receiving patient data; A step of providing data corresponding to the patient data and available operation parameters of the medical imaging device to one or more input nodes of the above-mentioned trained machine learning model; A step of receiving one or more recommended action parameter settings through one or more output nodes of the above-mentioned trained machine learning model; A step of transmitting data to the medical imaging device that causes modification of the operation parameters of the medical imaging device based on the above one or more recommended operation parameter settings; A step of receiving metadata corresponding to one or more images captured by the medical imaging device; and A method of using a machine learning technique comprising the step of further training the trained machine learning model based on the above metadata.
In Article 11, A method using a machine learning technique, further comprising the step of determining one or more additional modifications made to the operation parameters of the medical imaging device, and the step of further training the trained machine learning model is further based on the one or more additional modifications.
In Article 11, The step of receiving the metadata corresponding to one or more images captured by the medical imaging device is: A method using a machine learning technique comprising the step of determining the metadata by processing the one or more images to identify the image quality for each of the one or more images.
In Article 11, The step of transmitting data that causes modification of the operation parameters of the medical imaging device is: The above patient data; and A method using a machine learning technique comprising the step of determining the data based on one or more recommended operation parameter settings.
In Article 11, The step of receiving the above patient data is: A step of displaying a user interface through a user device; and A method using machine learning techniques comprising the step of receiving patient weight and patient height through the above user interface.
In Article 11, The step of determining available operating parameters of the medical imaging device is: A method using machine learning techniques comprising the step of querying a database of device identifications and corresponding operation parameters based on the indication of the medical imaging device.
In one or more non-transient computer-readable memories configured to use machine learning techniques to optimize medical image processing to reduce unintended harmful radiation exposure, The above memory stores instructions, and when the instructions are executed by one or more processors of a computing device, the computing device: A machine learning model is generated by training a machine learning model to output recommended medical imaging device operation parameter settings using training data including a history of medical imaging device operation parameter settings, patient data, and image results, and training the machine learning model includes modifying one or more weights of one or more nodes of an artificial neural network based on the training data; To receive the indication of the medical imaging device; Based on the display of the medical imaging device, the available operating parameters of the medical imaging device are determined; To receive patient data; To provide data corresponding to the patient data and available operating parameters of the medical imaging device to one or more input nodes of the above-mentioned trained machine learning model; Receiving one or more recommended action parameter settings through one or more output nodes of the above-mentioned trained machine learning model; Data that causes modification of the operating parameters of the medical imaging device based on one or more recommended operating parameter settings is transmitted to the medical imaging device; Receiving metadata corresponding to one or more images captured by the medical imaging device; One or more non-transient computer-readable memories that allow the trained machine learning model to be further trained based on the above metadata.
In Article 17, When the above instructions are executed by the above one or more processors, the computing device: One or more non-transient computer-readable memories that determine one or more additional modifications made to the operating parameters of the medical imaging device, and, when the instructions are executed by the one or more processors, cause the computing device to further train the trained machine learning model based on the one or more additional modifications.
In Article 17, One or more non-transient computer-readable memories, wherein the above instructions, when executed by one or more processors, cause the computing device to determine metadata by processing one or more images to identify image quality for each of one or more images, thereby causing the computing device to receive said metadata corresponding to one or more images captured by the medical imaging device.
In Article 17, When the above instructions are executed by one or more processors, the computing device: The above patient data; and Based on the above one or more recommended operation parameter settings, One or more non-transient computer-readable memories that cause the computing device to transmit data that causes a modification of the operating parameters of the medical imaging device by determining the above data.

Description

Medical imaging systems for reducing radiation exposure Aspects of the present disclosure generally relate to medical imaging systems such as fluoroscopic devices and X-ray devices. More specifically, aspects of the present disclosure may provide for reducing or eliminating high doses of harmful radiation by using a combination of patient-specific data and/or machine learning techniques. Medical imaging devices, such as fluoroscopic imaging devices and X-ray imaging devices, can be used to produce visual representations of the internal structures of the human body, animals, or other inanimate objects. While these imaging devices are very useful, they can emit radiation that is harmful to both the subject being imaged or scanned and nearby objects (e.g., radiographers, doctors, etc.). For example, typical medical imaging systems can extend exposure to more than 0.5 seconds, and radiation is detectable from more than 6 feet away with scattering rates of more than 300 millirem/hour (mRem/hour) or higher for all images generated during a given imaging session. Excessive exposure to such radiation can cause various diseases, such as cataracts and various forms of cancer. Incorrect configuration and/or misuse of medical imaging devices can increase the incidence of radiation exposure to the scanned subject and nearby objects. For example, some radiographers repeatedly use medical imaging devices in manual mode until the desired image is obtained—that is, the radiographer may use the medical imaging system to take a first image, evaluate image quality, modify the system's operating parameters (e.g., kVp, mAs), and then re-imaging. Most commonly, this process (repeated imaging, image quality evaluation, parameter modification, and re-imaging) is automated in a process known as Automated Exposure Control (AEC). AEC is a safety feature commonly used in more modern imaging devices to reduce radiation exposure. Nevertheless, every new image exposes the subject and nearby objects to additional radiation. While this may not be a significant issue for patients who are scanned only a few times in their lifetime, it can pose a significant risk to nearby physicians and radiographers who may be regularly exposed to such radiation. The following presents a simplified summary of the various embodiments described herein. This summary is not an extensive overview and is not intended to identify key or important elements or to limit the scope of the claims. The following summary merely presents some concepts in a simplified form as an introduction to the more detailed description provided below. The embodiments described herein relate to the use of machine learning to improve the process of configuring medical imaging systems, thereby reducing overall exposure to patients, clinical personnel, etc. To achieve this goal, the embodiments described herein relate to training a machine learning model based on training data including medical imaging device operation parameter settings, patient data, and the history of imaging results, and to developing a complete model of ideal parameter settings that considers variables such as patient height/weight, specific characteristics of a particular medical imaging system model, and operation parameters available in different medical imaging system models. This trained machine learning model may receive information regarding available operation parameters and patient data for a particular medical imaging system, and the trained machine learning model may then output recommended operation parameter settings. These recommended operation parameter settings, along with other considerations, may be used to generate guidelines for the medical imaging system. Metadata regarding the image(s) captured by the medical imaging system may then be used to further train the model, ensuring that the model learns iteratively over time. This process has numerous advantages revealed during testing. On the one hand, medical imaging systems are performed more accurately and quickly, reducing total radiation exposure to the subject and people around the subject, such as doctors and staff. On the other hand, this process also lowers the overall power consumption and radiation output of the entire medical imaging system, resulting in significant cost savings, improved system lifespan, and reduced likelihood of the medical imaging system becoming temporarily unusable due to overheating issues. More specifically, the computing device may generate a trained machine learning model by using training data that includes medical imaging device operation parameter settings, patient data, and a history of imaging results, and training the machine learning model to output recommended medical imaging device operation parameter settings. Training the machine learning model may include modifying one or more weights of one or more nodes of an artificial neural network based on the training data. The computing device may receive a dis