KR-20260065479-A - SYSTEM AND METHOD FOR SAGITTAL PLANE CORRECTION OF MEDICAL IMAGES

KR20260065479AKR 20260065479 AKR20260065479 AKR 20260065479AKR-20260065479-A

Abstract

The present invention includes a processing unit that reconstructs the shape of a spinal structure included in a medical image, wherein the processing unit inputs the medical image into a pre-trained deep learning model to segment the vertebral bodies, connects the center points of the segmented vertebral bodies to form a centerline, analyzes the curvature of the centerline to establish a sagittal plane reference correction coordinate system, transforms the shape of the spinal structure based on the correction coordinate system to generate a sagittal plane cross-sectional image, and outputs the sagittal plane cross-sectional image as a readout image.

Inventors

김종효
이제명
김민범

Assignees

주식회사 클라리파이

Dates

Publication Date: 20260508
Application Date: 20250527
Priority Date: 20241031

Claims (15)

It includes a processing unit that reconstructs the shape of spinal structures included in medical images, and The above processing unit is, The above medical image is input into a pre-trained deep learning model to segment the vertebral bodies; Construct a centerline by connecting the center points of the divided vertebrae; Analyze the curvature of the above centerline to establish a sagittal plane-based correction coordinate system; Generate a sagittal cross-sectional image by transforming the shape of the above-mentioned spinal structure based on the above-mentioned correction coordinate system; A sagittal plane correction device for medical images that outputs the above sagittal plane cross-sectional image as a readout image.
In paragraph 1, The above deep learning model is, A sagittal plane correction device for medical images characterized by performing segmentation of vertebral bodies based on a mask, including a U-NET or Attention structure.
In paragraph 2, The above processing unit is, A sagittal plane correction device for medical images, characterized by calculating coordinates for each of the above-mentioned vertebrae to derive individual center points for the above-mentioned vertebrae, and connecting the center points of the above-mentioned vertebrae to form a three-dimensional centerline.
In paragraph 3, The above processing unit is, A sagittal plane correction device for a medical image, characterized by applying at least one of linear interpolation, a cubic Bezier curve, and a cubic spline in the configuration of the above three-dimensional centerline.
In paragraph 3, The above processing unit is, A sagittal plane correction device for a medical image, characterized by defining a starting point and an ending point in the medical image as a straight line, determining the position between the two points, and then calculating an intermediate position to construct the 3D centerline.
In paragraph 3, The above processing unit is, A sagittal plane correction device for a medical image, characterized by defining a starting point and an ending point in the medical image and determining the shape of a curve through control points arranged in the direction of curvature or the direction of tangent to construct the three-dimensional centerline.
In paragraph 3, The above processing unit is, A sagittal plane correction device for a medical image, characterized by connecting the centerlines of each vertebral body in the medical image with curves and defining spline curves for each axis of the global coordinate system to construct the three-dimensional centerlines.
In paragraph 1, The above processing unit is, Calculate the tangent vector and curvature of a preset point in the curvature analysis of the above centerline, and The above-mentioned preset point is, A sagittal plane correction device for a medical image, characterized by including at least one of a center point of a vertebral body and an interpolated point between vertebral bodies.
In paragraph 8, The above processing unit is, A sagittal plane correction device for medical images, characterized by calculating the direction of progression of the centerline based on the difference between the center points of the vertebral bodies in the calculation of the above tangent vector.
In paragraph 8, The above processing unit is, A sagittal plane correction device for medical images, characterized by calculating the degree of curvature of the centerline based on the difference in center points between vertebral bodies in the calculation of the above curvature.
In paragraph 8, The above processing unit is, A sagittal plane correction device for a medical image, characterized by using a local coordinate system defined based on the tangent vector and curvature of the centerline in the setting of the correction coordinate system.
In Paragraph 11, The above processing unit is, A sagittal plane correction device for a medical image, characterized by transforming the medical image according to a correction coordinate system and reconstructing the medical image into a sagittal plane cross-sectional image with curvature removed.
In paragraph 1, The medical image above is, A sagittal plane correction device for medical images characterized by including computed tomography images of the thoracic or lumbar region.
A communication unit for acquiring medical images; and It includes a processing unit that reconstructs the shape of the spinal structure included in the medical image above, and The above processing unit is, The above medical image is input into a pre-trained deep learning model to segment the vertebral bodies; Construct a centerline by connecting the center points of the divided vertebrae; Analyze the curvature of the above centerline to establish a sagittal plane-based correction coordinate system; Generate a sagittal cross-sectional image by transforming the shape of the above-mentioned spinal structure based on the above-mentioned correction coordinate system; A sagittal plane correction system for medical images that outputs the above sagittal plane cross-sectional image as a readout image.
In a sagittal plane correction method for medical images that reconstructs the shape of spinal structures included in the medical image, A step of segmenting the vertebral bodies by inputting the above medical image into a pre-trained deep learning model; A step of forming a centerline by connecting the center points of the divided vertebrae; A step of establishing a sagittal plane-based correction coordinate system by analyzing the curvature of the above centerline; A step of generating a sagittal cross-sectional image by transforming the shape of the above-mentioned spine structure based on the above-mentioned correction coordinate system; and A method for correcting the sagittal plane of a medical image, comprising the step of outputting the above sagittal plane cross-sectional image as a readout image.

Description

System and Method for Sagittal Plane Correction of Medical Images The present invention relates to a system and method for correcting the sagittal plane of a medical image, and more specifically, to a system and method for correcting the spinal structure of a patient with scoliosis in a medical image to reconstruct a distorted sagittal plane cross-section into an aligned state. Generally, Curved Planar Reformation (CPR) techniques are used in the diagnosis and evaluation of scoliosis. Computerized Photometric Radiography utilizes computer-based optical photometry technology to enable the precise diagnosis and evaluation of a patient's spinal deformity. Computerized Photometric Radiography is non-invasive and is useful for the diagnosis and tracking of scoliosis because it allows for 3D modeling and accurate angle measurement through computer analysis of captured images. In this computerized photometric radiography technique, the three-dimensional curvature of the spine is measured by manually specifying the center point of the vertebral body and generating a curved path. However, in the diagnosis of spinal diseases, the diagnostic time is long because multiple slices are analyzed manually, and it is difficult to identify the entire length of the spine at once, which limits the diagnostic efficiency and accuracy. FIG. 1 is a conceptual diagram showing a sagittal plane correction system for medical images according to the present embodiment, and FIG. 2 is a flowchart illustrating a sagittal plane correction method using a sagittal plane correction system for medical images according to the present embodiment, and FIG. 3 is a conceptual diagram illustrating a sagittal plane correction method using a sagittal plane correction program installed in a sagittal plane correction system for medical images according to the present embodiment, and FIG. 4 is a diagram showing a medical image and a reading image input to a sagittal plane correction system for medical images according to the present embodiment. Embodiments of the present invention will be described in detail below with reference to the attached drawings. However, the embodiments disclosed below are not limited to those disclosed below and may be implemented in various forms; the embodiments provided are merely intended to ensure that the disclosure of the present invention is complete and to fully inform those skilled in the art of the scope of the invention. The shapes of elements in the drawings may be exaggerated for clearer explanation, and elements indicated by the same reference numeral in the drawings represent the same element. FIG. 1 is a conceptual diagram showing a sagittal plane correction system for medical images according to the present embodiment. As illustrated in FIG. 1, the sagittal plane correction system for medical images according to the present embodiment (1000, hereinafter referred to as the correction system) includes a sagittal plane correction device (100). The sagittal plane correction device (100) corrects the spinal structure of a patient with a spinal disease in a medical image (10) obtained from a medical imaging device, and reconstructs a distorted sagittal plane cross-section into an aligned state. The medical image (10) may be a computed tomography image of the thoracic or lumbar region, but is not limited thereto. The sagittal plane correction device (100) is composed of an electronic device in which a sagittal plane correction program (P) can be installed, and various electronic devices including display elements such as PCs, netbooks, tablet PCs, and smartphones may be used. Meanwhile, the sagittal plane correction device (100) includes a communication unit (110), a data storage unit (120), and a processing unit (130). The communication unit (110) receives a medical image (10) from a medical image capturing device or a separate storage device. The data storage unit (120) is configured to include memory and stores a sagittal plane correction program (P). The sagittal plane correction program (P) can reconstruct the sagittal plane of the medical image (10) and output an improved reading image (30) to the medical staff. The processing unit (130) performs overall control of the sagittal plane correction device (100) and can create and execute a process for sagittal plane correction work based on the sagittal plane correction program (P). Hereinafter, a sagittal plane correction method using a correction system (1000) according to the present embodiment will be described in detail with reference to the attached drawings. FIG. 2 is a flowchart illustrating a sagittal plane correction method using a sagittal plane correction system for medical images according to the present embodiment, and FIG. 3 is a conceptual diagram illustrating a sagittal plane correction method using a sagittal plane correction program loaded into a sagittal plane correction system for medical images according to the present embodiment. FIG. 4 is a d