CN-122006149-A - Image processing-based neck radiotherapy method for nasopharyngeal carcinoma patients

CN122006149ACN 122006149 ACN122006149 ACN 122006149ACN-122006149-A

Abstract

The invention discloses a nasopharyngeal carcinoma patient neck radiotherapy method based on image processing, which relates to the technical field of medical image processing and comprises the following steps: collecting a neck original image containing metal interference, drawing a gray mutation distribution diagram around a high-density abnormal region, and extracting continuous stripe trend according to the gray mutation distribution diagram to form an interference profile set. According to the invention, by constructing the interference contour set and the artifact removal region diagram, a space limit range is established in advance, high-density artifacts are blocked from participating in tumor boundary generation, the target region delineating accuracy is improved, and meanwhile, an expansion stopping and rollback control mechanism is introduced in contour smoothing and adjacent slice comparison, so that false expansion in the three-dimensional direction is inhibited, the irradiation risk of normal tissues is reduced, and the radiotherapy safety and boundary stability are improved.

Inventors

JIANG MINGHUA
YU LE
JIANG JING

Assignees

江苏省肿瘤医院

Dates

Publication Date: 20260512
Application Date: 20260410

Claims (8)

1. The nasopharyngeal carcinoma patient neck radiotherapy method based on image processing is characterized by comprising the following steps of: collecting a neck original image containing metal interference, drawing a gray mutation distribution diagram around a high-density abnormal region, and extracting continuous stripe trend according to the gray mutation distribution diagram to form an interference profile set; Continuously tracking along the continuous stripe trend in the neck original image by utilizing the interference contour set, separating the artifact areas which are radially diffused, generating an artifact removal area map, and covering the artifact removal area map to the neck original image to form a boundary limit range; Gray scale decreasing screening is carried out on the high-density area outside the boundary limit range, and the focus edge which is complete in closing and stable in gray scale distribution is reserved, so that a limited tumor boundary diagram is formed; Performing contour smoothing around the limited tumor boundary map, performing space coordinate comparison on each outward-expansion pixel in the contour smoothing process, and stopping the outward-expansion operation in the current direction when the outward-expansion pixel is positioned in the coverage range of the artifact removal region map to form a controlled tumor contour result; And comparing the adjacent sections around the controlled tumor contour result, and when the controlled tumor contour result has sudden jump, backing the controlled tumor contour result to the corresponding range of the limited tumor boundary map to complete the continuous suppression of false expansion.
2. The image processing-based method for neck radiotherapy of a patient with nasopharyngeal carcinoma according to claim 1, wherein the interference profile set forming step is as follows: pixel-by-pixel gray level analysis is carried out on the neck original image, all pixels are traversed according to a line sequence and a sequence, gray level difference values between each pixel and eight adjacent pixels in the direction are calculated, pixels with the gray level difference values exceeding a preset gradient threshold value are marked as gray level mutation points, and all gray level mutation points are collected to generate a gray level mutation distribution map; Carrying out space aggregation analysis on gray abrupt change points in a gray abrupt change distribution map to form gray abrupt change aggregation areas, extracting gray abrupt change points with consistent directions from each gray abrupt change aggregation area, carrying out extension treatment, integrating extension tracks with included angles of the directions being smaller than a set angle range, and forming a continuous stripe trend set; mapping the continuous stripe trend set to the neck original image, establishing a stripe expansion band along the continuous stripe trend set and forming a stripe expansion area, and carrying out space combination and communication treatment on the stripe expansion area to form an interference contour area in a surrounding manner; and carrying out boundary correction and region integration on the interference contour region to form an interference contour set covering the high-density abnormal region.
3. The image processing-based method of neck radiotherapy for a patient with nasopharyngeal carcinoma according to claim 2, wherein the steps of generating an artifact removal region map and forming a boundary limit range are as follows: Mapping the interference contour set to the neck original image, extracting the pixel point coordinates of each contour unit boundary, advancing point by point along the continuous stripe trend, and forming a continuous tracking path set according to the gray difference value and the gray change direction; Grouping the continuous tracking path sets according to the initial contour units, and combining the continuous tracking paths which are adjacent in angle and have the space interval within a set pixel distance range to form an initial artifact diffusion region set; Boundary scanning and communication processing are carried out on the initial artifact diffusion region set, and an artifact removal region diagram is formed by enclosing; Covering the artifact removal region diagram to the neck original image, marking pixels in the artifact removal region diagram, and constructing a boundary limit range.
4. The image processing-based method for neck radiotherapy of a patient with nasopharyngeal carcinoma according to claim 3, wherein the step of forming the restricted tumor boundary map is as follows: expressing the boundary limit range in a pixel mark form, reading gray values point by point outside the boundary limit range, dividing gray levels according to the descending order of the gray values, and performing space communication processing to form a candidate high-density region set; Carrying out gray level distribution stability judgment on the candidate high-density region set, constructing a gray level change path along the horizontal, vertical and diagonal directions, and reserving candidate high-density regions with gray level decreasing continuity to form an initial focus edge candidate set; Performing closed integrity processing on the initial focus edge candidate set, extracting boundary pixels and connecting to form a closed path, filling the inside of the closed path and correcting gray scale inverse distribution pixels; and integrating and expressing the regions meeting the conditions of complete closure and stable gray distribution to form a limited tumor boundary diagram.
5. The method according to claim 4, wherein when the gray level distribution stability is determined, the candidate high-density regions are continuously read along a gray level change path from the center to the boundary direction for each candidate high-density region, and when the gray level change path has inverse distribution pixels, the corresponding regions are removed, and the boundary pixels of the remaining regions are subjected to reconnection correction to define a restricted tumor boundary map which only includes regions with continuously decreasing gray levels and closed boundaries.
6. The image processing-based method of neck radiotherapy in a patient with nasopharyngeal carcinoma of claim 4, wherein the step of forming the controlled tumor contour results is as follows: reading boundary pixel coordinates of the limited tumor boundary map point by point, and arranging boundary pixels into a continuous closed boundary point sequence according to a space sequence; Performing curvature analysis on each boundary pixel in the boundary point sequence, and inserting a new pixel point to convert the broken line form into a continuous curve form so as to ensure that the gray value is consistent with the limited tumor boundary map; performing outward expansion processing on the smoothed boundary set, advancing along the normal direction of the boundary, comparing whether the artifact is entering the artifact removal region map in real time, stopping outward expansion when the artifact removal region map is coincident, and reserving the last pixel point which is not entering the artifact removal region map; Connectivity trimming is carried out on the updated boundary set, new boundary pixel points are inserted, the concave or convex boundary sections are corrected, the smooth boundary curve is ensured to be continuous in space, and finally a controlled tumor contour result is formed.
7. The image processing-based method for neck radiotherapy of a nasopharyngeal carcinoma patient according to claim 6, wherein during the outward expansion processing, the boundary pixels are synchronously aligned with each other in the normal direction, and the pushing path and the boundary of the artifact removal region map are spatially coincident, i.e. the pushing is terminated, and the corresponding boundary pixels are defined within the range of the limited tumor boundary map.
8. The image processing-based method for neck radiotherapy of a patient with nasopharyngeal carcinoma of claim 6, wherein the step of comparing adjacent slices back and forth and controlling back is as follows: Numbering all the slices according to the acquisition sequence, expressing the controlled tumor contour result and the limited tumor boundary map of each numbered slice in a unified space coordinate system, selecting a target slice and establishing a space corresponding relation with the previous numbered slice and the next numbered slice; carrying out area statistics and boundary displacement analysis on the controlled tumor contour result of the target slice, identifying boundary pixels with the spatial distance exceeding a set pixel displacement threshold value, and forming a burst outer jump region set; Replacing boundary pixel coordinates in the burst jumping region set with boundary pixel coordinates in a range corresponding to the limited tumor boundary map, updating the boundary pixel set and carrying out communication processing; And (3) performing spatial comparison on the updated controlled tumor contour result and the adjacent slices again, and repeating the area statistics and the boundary displacement analysis to complete the continuous suppression of false expansion.

Description

Image processing-based neck radiotherapy method for nasopharyngeal carcinoma patients Technical Field The invention relates to the technical field of medical image processing, in particular to a nasopharyngeal carcinoma patient neck radiotherapy method based on image processing. Background The image processing-based nasopharyngeal carcinoma patient neck radiotherapy is to acquire image data of the patient neck through medical image equipment, such as tomographic images from CT scanning, MRI inspection or PET-CT inspection, and then accurately identify and position the primary tumor range, the affected neck lymph node range and surrounding important organs by using image processing technologies such as image segmentation, registration, enhancement and boundary identification, and determine the range of a radiotherapy target area, dose distribution and irradiation path on the basis, thereby reducing the radiation damage to normal tissues such as spinal cord, parotid gland and throat as much as possible while ensuring that the tumor tissues obtain enough irradiation dose, and realizing an individual neck radiotherapy method taking medical image data as decision basis and taking accurate control as a target. The prior art has the following defects: In the prior art, when the image artifact removal treatment process before the neck radiotherapy of a nasopharyngeal carcinoma patient is carried out aiming at the high-density interference areas such as metal dental crowns and the like, if the artifact removal parameter is improperly set or the boundary discrimination rule is not fine enough, the residual high-density artifact is easily identified as a tumor edge by mistake, and the residual high-density artifact is automatically expanded in the subsequent contour smoothing and continuity filling process, so that a false tumor contour exceeding the range of a real focus is formed. The false contour can directly participate in drawing and calculating the dose distribution of a radiotherapy target area, so that the irradiation range is intangibly enlarged to the surrounding normal neck muscle area, muscle fibrosis and function limitation are easily caused, irreversible function damage such as dysphagia is caused when the false contour is serious, and great clinical potential safety hazard exists. The above information disclosed in the background section is only for enhancement of understanding of the background of the disclosure and therefore it may include information that does not form the prior art that is already known to a person of ordinary skill in the art. Disclosure of Invention The invention aims to provide a nasopharyngeal carcinoma patient neck radiotherapy method based on image processing, so as to solve the problems in the background technology. In order to achieve the purpose, the invention provides the technical scheme that the method for radiotherapy of the neck of a nasopharyngeal carcinoma patient based on image processing comprises the following steps: collecting a neck original image containing metal interference, drawing a gray mutation distribution diagram around a high-density abnormal region, and extracting continuous stripe trend according to the gray mutation distribution diagram to form an interference profile set; Continuously tracking along the continuous stripe trend in the neck original image by utilizing the interference contour set, separating the artifact areas which are radially diffused, generating an artifact removal area map, and covering the artifact removal area map to the neck original image to form a boundary limit range; Gray scale decreasing screening is carried out on the high-density area outside the boundary limit range, and the focus edge which is complete in closing and stable in gray scale distribution is reserved, so that a limited tumor boundary diagram is formed; Performing contour smoothing around the limited tumor boundary map, performing space coordinate comparison on each outward-expansion pixel in the contour smoothing process, and stopping the outward-expansion operation in the current direction when the outward-expansion pixel is positioned in the coverage range of the artifact removal region map to form a controlled tumor contour result; And comparing the adjacent sections around the controlled tumor contour result, and when the controlled tumor contour result has sudden jump, backing the controlled tumor contour result to the corresponding range of the limited tumor boundary map to complete the continuous suppression of false expansion. Preferably, the interference profile set forming step is as follows: pixel-by-pixel gray level analysis is carried out on the neck original image, all pixels are traversed according to a line sequence and a sequence, gray level difference values between each pixel and eight adjacent pixels in the direction are calculated, pixels with the gray level difference values exceeding a preset gradient threshold value are marked as