Intelligent super-resolution and contrast-agent-free enhanced intracranial vascular wall MR image generation system

Abstract

The invention discloses an intelligent super-resolution and contrast-agent-free enhanced intracranial vascular wall MR image generation system, and belongs to the technical field of medical informatization, and the system comprises an image input module which employs an MRI image of a sagittal view 3D-TI-SPACE sequence after plain scanning and enhancement as input data, and guarantees that the image quality meets the requirements of subsequent image optimization and generation; the super-resolution plain scanning image generation module is used for generating a super-resolution AI-NC-hrVWI image based on the NC-hrVWI image by using a super-resolution model; and a synthetic contrast-agent-free enhanced image generation module. The intelligent super-resolution and contrast-agent-free enhanced intracranial vascular wall MR image generation system is a promising auxiliary tool for image optimization and generation in imaging examination, the artificial intelligence auxiliary advantage is utilized, the hrVWI examination image quality is improved, the hrVWI examination time is shortened, contrast agents are prevented from being used, the examination safety and universality are enhanced, and the system is suitable for being used in the field of imaging examination. Powerful assistance is better provided for clinical diagnosis and treatment work of the intracranial atherosclerosis disease through the hrVWI examination.

Description

Technical Field

[0001] The present invention belongs to the technical field of medical informatization, especially an intelligent super-resolution and non-contrast-enhanced intracranial vessel wall MR image generation system, and also relates to an operation method of the intelligent super-resolution and non-contrast-enhanced intracranial vessel wall MR image generation system. Background Art

[0002] Traditional lumen imaging technology has limitations in analyzing intracranial vascular lesions as it only evaluates changes in the vascular lumen and cannot detect the vessel wall. However, hrVWI is a technology that can directly detect pathological changes in the vessel wall in vitro and non-invasively, and evaluate the characteristics and vulnerability of atherosclerotic plaques.

[0003] Currently, high-resolution magnetic resonance vessel wall imaging mainly has two problems. Firstly, limited by the current software and hardware technology levels and the requirements of high-resolution image acquisition, the examination time of hrVWI is too long, and some patients cannot tolerate it. Secondly, hrVWI requires injection of gadolinium contrast agent for delayed enhancement scanning, and the vascular lesions are evaluated by combining non-enhanced and enhanced images. Although gadolinium contrast agent is generally safe, previous studies have shown that there are still risks of adverse reactions such as allergy and nephrogenic systemic fibrosis when injecting gadolinium contrast agent. In addition, patients with renal insufficiency, advanced age, allergic constitution, and asthma cannot use gadolinium contrast agent for enhanced examination. For the above reasons, the application of hrVWI technology in clinical practice is restricted to a certain extent.

[0004] Therefore, there is an urgent need to develop a model tool for image optimization and generation in high-resolution magnetic resonance intracranial vessel wall imaging, which simplifies the examination process, shortens the examination time, reduces the potential risks of using contrast agent, and has important clinical value for improving the safety and universality of this examination in clinical applications. Summary of the Invention

[0005] The purpose of the present invention is to use the artificial intelligence deep learning method to generate a super-resolution non-enhanced magnetic resonance vessel wall imaging and a synthesized non-contrast-enhanced high-resolution vessel wall imaging enhanced image based on the non-enhanced image of high-resolution magnetic resonance vessel wall imaging, so as to improve the image quality, shorten the examination time, reduce the potential allergy risk of using contrast agent, and improve the safety and universality of hrVWI examination.

[0006] To achieve the above purpose, the present invention provides the following technical solution: An intelligent super-resolution and non-contrast-enhanced intracranial vessel wall MR image generation system, including:

[0007] An image input module: uses the MRI images of the non-enhanced and enhanced sagittal 3D-TI-SPACE sequence as input data to ensure that the image quality meets the requirements of subsequent image optimization and generation;

[0008] Ultra-high resolution non-contrast enhanced image generation module: Using an ultra-high resolution model, an AI-NC-hrVWI image with ultra-high resolution is generated based on the NC-hrVWI image;

[0009] Contrast agent-free enhanced image generation module: Using an image generation model and an ultra-high resolution model, a Syn-CE-hrVWI image is generated and super-resolved based on the NC-hrVWI image, and the resolution of the Syn-CE-hrVWI image is kept consistent with that of the AI-NC-hrVWI image;

[0010] Image evaluation module: The quality of the synthesized virtual images is evaluated from three dimensions: quantitative, visual, and diagnostic. At the same time, professional doctors evaluate the diagnostic efficacy of the generated images to ensure the usability of the generated images in medical diagnosis;

[0011] Model fusion module: The trained ultra-high resolution model and the image generation model are fused to construct an artificial intelligence image optimization and generation joint model, realizing the function of one-stop output of AI-NC-hrVWI and Syn-CE-hrVWI images based on the NC-hrVWI image.

[0012] As a preferred implementation, the non-contrast enhanced method is set to NC-hrVWI, and the enhanced method is set to CE-hrVWI.

[0013] As a preferred implementation, the quantitative evaluation is used to measure the similarity between the generated image and the real image.

[0014] As a preferred implementation, the visual evaluation is used for professional doctors or evaluators to give subjective judgments on the authenticity and clarity of the generated images.

[0015] Operation method of an intelligent ultra-high resolution and contrast agent-free enhanced intracranial vessel wall MR image generation system, including the following steps:

[0016] Step 1: Import the MRI images of the sagittal 3D-TI-SPACE sequences of the non-contrast enhanced and enhanced intracranial vessel walls of the patient;

[0017] Step 2: The NC-hrVWI image outputs the AI-NC-hrVWI and Syn-CE-hrVWI images in one-stop based on the artificial intelligence image optimization and generation joint model;

[0018] Step 3: Evaluate the quality of the synthesized virtual images from three dimensions: quantitative, visual, and diagnostic. The system quantitatively evaluates the similarity between the generated image and the real image, and professional doctors or evaluators evaluate the authenticity, clarity, and diagnostic efficacy of the generated images to ensure the usability of the generated images in medical diagnosis.

[0019] Technical effects and advantages of the present invention compared with the prior art:

[0020] The intelligent super-resolution and non-contrast-enhanced intracranial vessel wall MR image generation system improves the image quality of hrVWI examinations: By constructing a super-resolution model of hrVWI through artificial intelligence deep learning technology, it provides high-quality images for clinical work that can be used for interpretation and clinical practice.

[0021] The intelligent super-resolution and non-contrast-enhanced intracranial vessel wall MR image generation system shortens the overall time of hrVWI examinations, avoids the use of contrast agents, and enhances the safety and universality of the examinations: By realizing the generation from NC-hrVWI to Syn-CE-hrVWI, it directly shortens the examination time. At the same time, it can avoid the use of gadolinium contrast agents, effectively improve the examination safety, and solve the limitations of patients who cannot use gadolinium contrast agents for enhanced examinations in hrVWI examinations, broaden the applicable population of hrVWI examinations, improve the universality of this examination, and better assist clinical evaluation, diagnosis, and treatment of intracranial artery-related diseases.

[0022] The intelligent super-resolution and non-contrast-enhanced intracranial vessel wall MR image generation system promotes scientific research development: It provides reliable imaging data support for the scientific research of intracranial atherosclerosis diseases, and helps to promote the medical-industrial cross-integration development in the field of vessel wall research and the field of artificial intelligence research.

[0023] The intelligent super-resolution and non-contrast-enhanced intracranial vessel wall MR image generation system is a promising auxiliary tool for image optimization and generation in imaging examinations. It utilizes the advantages of artificial intelligence assistance to improve the image quality of hrVWI examinations, shorten the time of hrVWI examinations, avoid the use of contrast agents, enhance the safety and universality of the examinations, and better enable hrVWI examinations to provide strong assistance for the diagnosis and treatment of clinical intracranial atherosclerosis diseases. Brief Description of the Drawings

[0024] Figure 1 is the framework diagram of the present invention;

[0025] Figure 2 is the flowchart of the present invention. Detailed Embodiments

[0026] In the following description, a large number of specific details are given to provide a more thorough understanding of the present invention. However, it is obvious to those skilled in the art that the present invention can be implemented without one or more of these details. In other examples, in order to avoid confusion with the present invention, some technical features well known in the art are not described.

[0027] Please refer to Figures 1 to 2, an intelligent super-resolution and contrast agent-free enhanced intracranial vessel wall MR image generation system. In this embodiment, it includes:

[0028] Image input module: Using the MRI images of the sagittal 3D-TI-SPACE sequence in non-contrast-enhanced (NC) and contrast-enhanced (CE) scans as input data to ensure that the image quality meets the requirements for subsequent image optimization and generation.

[0029] Super-resolution non-contrast-enhanced image generation module: Using a super-resolution model to generate super-resolution AI-NC-hrVWI images based on NC-hrVWI images.

[0030] Synthetic contrast agent-free enhanced image generation module: Using an image generation model and a super-resolution model to generate and super-resolve Syn-CE-hrVWI images based on NC-hrVWI images, and keeping the resolution of Syn-CE-hrVWI images consistent with that of AI-NC-hrVWI images.

[0031] Image evaluation module: Evaluating the quality of the synthesized virtual images from three dimensions: quantitative, visual, and diagnostic. At the same time, professional doctors evaluate the diagnostic efficacy of the generated images to ensure the usability of the generated images in medical diagnosis.

[0032] Model fusion module: Fusing the trained super-resolution model and the image generation model to construct an artificial intelligence image optimization and generation joint model, and realizing the function of one-stop output of AI-NC-hrVWI and Syn-CE-hrVWI images based on NC-hrVWI images.

[0033] The non-contrast-enhanced scan method is set as NC-hrVWI, and the contrast-enhanced scan method is set as CE-hrVWI.

[0034] The quantitative evaluation is used to measure the similarity between the generated images and the real images.

[0035] The visual evaluation is used for professional doctors or evaluators to give subjective judgments on the authenticity and clarity of the generated images.

[0036] The operation method of the intelligent super-resolution and contrast agent-free enhanced intracranial vessel wall MR image generation system includes the following steps:

[0037] Step 1: Import the MRI images of the sagittal 3D-TI-SPACE sequence in non-contrast-enhanced and contrast-enhanced scans of the patient's intracranial vessel wall.

[0038] Step 2: The NC-hrVWI images output AI-NC-hrVWI and Syn-CE-hrVWI images in one-stop based on the artificial intelligence image optimization and generation joint model.

[0039] Step 3: Evaluate the quality of the synthesized virtual image from three dimensions: quantification, visualization, and diagnosis. The system quantitatively evaluates the similarity between the generated image and the real image, and professional doctors or evaluators evaluate the authenticity, clarity, and diagnostic efficacy of the generated image to ensure the usability of the generated image in medical diagnosis.

[0040] It should be noted that in this article, relational terms such as "one" and "two" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations.

[0041] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent super-resolution and non-contrast agent enhanced intracranial vessel wall MR image generation system, characterized in that: Including: Image input module: Using the MRI images of the sagittal 3D-TI-SPACE sequence in non-contrast-enhanced and contrast-enhanced scans as input data to ensure that the image quality meets the requirements of subsequent image optimization and generation. Super-resolution non-contrast-enhanced image generation module: Using a super-resolution model to generate super-resolution AI-NC-hrVWI images based on NC-hrVWI images. Synthetic contrast-agent-free enhanced image generation module: Using an image generation model and a super-resolution model to generate and super-resolve output Syn-CE-hrVWI images based on NC-hrVWI images, keeping the resolution of Syn-CE-hrVWI images consistent with that of AI-NC-hrVWI images. Image evaluation module: Evaluating the quality of the synthesized virtual images from three dimensions: quantitative, visual, and diagnostic. At the same time, professional doctors evaluate the diagnostic efficacy of the generated images to ensure the usability of the generated images in medical diagnosis. Model fusion module: Fusing the trained super-resolution model and the image generation model to construct an artificial intelligence image optimization and generation joint model, realizing the function of one-stop output of AI-NC-hrVWI and Syn-CE-hrVWI images based on NC-hrVWI images.

2. The intelligent super-resolution and non-contrast agent enhanced intracranial vessel wall MR image generation system according to claim 1, wherein: The non-contrast-enhanced scan method is set as NC-hrVWI, and the contrast-enhanced scan method is set as CE-hrVWI.

3. The intelligent super-resolution and non-contrast-enhanced intracranial vessel wall MR image generation system according to claim 1, characterized in that: The quantitative evaluation is used to measure the similarity between the generated images and the real images.

4. The intelligent super-resolution and contrast-agent-free enhanced intracranial vessel wall MR image generation system according to claim 1, wherein: The visual evaluation is used for professional doctors or evaluators to give subjective judgments on the authenticity and clarity of the generated images.

5. Method for operating an intelligent super-resolution and non-contrast agent enhanced intracranial vessel wall MR image generation system, for operating and implementing an intelligent super-resolution and non-contrast agent enhanced intracranial vessel wall MR image generation system, characterized in that: Including the following steps: Step 1: Import the MRI images of the sagittal 3D-TI-SPACE sequence in non-contrast-enhanced and contrast-enhanced scans of the patient's intracranial vessel wall. Step 2: The NC-hrVWI images output AI-NC-hrVWI and Syn-CE-hrVWI images in one stop based on the artificial intelligence image optimization and generation joint model. Step 3: Evaluate the quality of the synthesized virtual images from three dimensions: quantitative, visual, and diagnostic. The system quantitatively evaluates the similarity between the generated images and the real images, and professional doctors or evaluators evaluate the authenticity, clarity, and diagnostic efficacy of the generated images to ensure the usability of the generated images in medical diagnosis.

Images