CN-121998958-A - Orthopedics patient wound nursing method based on visual image processing

CN121998958ACN 121998958 ACN121998958 ACN 121998958ACN-121998958-A

Abstract

The invention discloses a visual image processing-based orthopedic patient wound care method, which relates to the technical field of wound care and comprises the following steps of constructing a patient information database based on patient basic information for an orthopedic patient, recording the patient basic information and wound information, acquiring daily images of the orthopedic patient after suturing the wound through an image acquisition device and recording the daily images in the patient information database, extracting characteristic pixels of the wound images in the patient information database and projecting the characteristic pixels into blank images, calibrating initial wound line pixels and initial line pixels in the initial images, numbering line heads, recording the initial line head line pixels and the pixel quantity, and the initial wound line pixels and the pixel quantity. The invention monitors the wound healing state through visual image processing, dynamically judges the disconnecting time and adapts to the individual difference of the patients, thereby improving the wound nursing efficiency of the orthopedic patients.

Inventors

LI LIHUA
LI JINJU
YAN XIAOXIA
GUO HUIMIN

Assignees

河南省洛阳正骨医院（河南省骨科医院）

Dates

Publication Date: 20260508
Application Date: 20260130

Claims (8)

1. A method of orthopedic patient wound care based on visual image processing, the method comprising the steps of: S1, constructing a patient information database aiming at an orthopedics patient based on patient basic information, recording the patient basic information and wound information, acquiring daily images of the orthopedics patient after wound suturing through an image acquisition device, and recording the daily images in the patient information database; s2, extracting characteristic pixels of a wound image aiming at the wound image in the patient information database, projecting the characteristic pixels into a blank image, calibrating initial wound line pixels and initial line pixels aiming at the initial image, numbering the line heads, and recording initial line pixels and pixel numbers, initial wound line pixels and pixel numbers; S3, defining a peripheral skin tissue area in an initial line of the initial image, and marking the number of the detachable line head every day according to the number of wound lines of the area of the line head with different numbers on the basis of the number of the pixels of the initial line head and the number of the pixels of the peripheral tissue, and judging the embedded degree of the line head with different numbers by combining the ratio change rates of the line heads with different numbers in the subsequent shooting image.
2. A method of orthopedic patient wound care based on visual image processing according to claim 1, wherein S1 comprises the steps of: S11, collecting basic information and wound information of an orthopedic patient, including patient name, contact information, age, wound position, wound length, operation date, stitching date and predicted stitch removal date, establishing a patient information database through MySQL, and establishing a file according to the patient name to record the collected information; S12, acquiring daily images of the wound suture position of the patient through the image acquisition equipment, and uploading the remarking date of the images to the corresponding patient file in the patient information database.
3. A method of orthopedic patient wound care based on visual image processing according to claim 1, wherein S2 comprises the steps of: S21, extracting a first patient wound image as an initial image through a current patient file in a patient information database, and identifying a wound and a suture head through an edge detection algorithm; S22, projecting the identified wound lines and the identified line head lines into blank images, recording an initial wound line pixel set and an initial line head line pixel set according to the projected initial images, numbering according to line heads of the current patient wound in sequence, and recording in corresponding patient files.
4. A method of orthopedic patient wound care based on visual image processing according to claim 3, characterized in that S21 comprises the steps of: S211, marking a first patient wound image in a current patient file in a patient information database as an initial image, performing color space conversion, and reducing skin texture noise through Gaussian smoothing filtering to obtain a preprocessed initial image I; s212, aiming at the preprocessed initial image I, calculating gradients in the horizontal and vertical directions of the initial image I through a Sobel operator, keeping local maxima through non-maxima suppression, refining edges, identifying edge points of the initial image through double-threshold processing, and outputting an edge map ; S213, aiming at the obtained edge graph Extracting all contours to obtain a set of contours Wherein each profile Consisting of successive edge pixels, the contour is determined to identify the wound and the suture head, comprising the steps of: Calculation of contours by the sum of Euclidean distances Distance of middle edge pixels to obtain length Calculating the minimum external rectangular length-width ratio Wherein Calculating curvature characteristics for the width and height of the rectangle Wherein Representing the curvature of each point on the contour; When (when) And is also provided with When in use, then Marked as candidate wound edges, wherein Respectively representing a length threshold value and a curvature high threshold value; When (when) And is also provided with Or (b) And (2) and When in use, then Marked as candidate suture head, wherein Representing a short length threshold, an aspect ratio threshold, and a low curvature threshold, respectively; S214, regarding each edge pixel point of the candidate wound edge on the outline, taking a neighborhood with the width of d at two sides along the normal direction, converting the two-side areas into HSV color space, extracting histograms p and q of H channels, and calculating the difference of the histograms by using the Pasteur distance When (when) If the current contour is larger than the difference threshold value, confirming that the current contour is a wound edge; For the minimum circumscribed rectangular area corresponding to the candidate suture thread head extraction outline, calculating the average saturation of the area in the HSV color space Wherein For the saturation value of pixel k, N is the number of regional pixels, and the direction variance is calculated Wherein Respectively representing the gradient direction and the average direction of the region when Less than the saturation threshold and If the current contour is smaller than the direction threshold value, the current contour is confirmed to be the suture thread head.
5. The method of orthopedic patient wound care based on visual image processing of claim 4, wherein S22 comprises the steps of: S221, adding all pixel points with the current outline confirmed as the wound edge into a wound edge mask Adding all pixel points with current outline being confirmed as suture head into suture head mask Creating a solid image of the same size as the original captured image, masking 、 All the pixel points 1 in the image are projected in pure color images in different colors; S222, recording a wound edge mask obtained from an initial image All 1 pixel points are initial wound line pixel sets, and a suture line head mask obtained by recording an initial image All 1 pixel points in the array are initial line head line pixel sets, and the array is aligned And (3) carrying out analysis of the communication component, separating and numbering the pixel areas which are communicated in space sequentially, and recording the pixel areas in the corresponding patient files.
6. A method of orthopedic patient wound care based on visual image processing according to claim 5, wherein S3 comprises the steps of: S31, defining a peripheral skin tissue area in an initial line head line of the initial image, taking each pixel of the line head line as a center, expanding the fixed pixel distance outwards and removing the line head pixels to obtain a peripheral region pixel set as a peripheral tissue pixel point; S32, judging the embedded degree of the thread ends based on the ratio of the number of the initial thread end pixel points to the number of the peripheral tissue pixel points and the ratio change rate of the thread ends with different numbers in the subsequent photographed image, and simultaneously marking the detachable thread end numbers every day according to the number of the daily wound line pixel concentrated pixel points in different thread end areas and the number of the initial wound line pixel concentrated pixel points.
7. The method of orthopedic patient wound care based on visual image processing of claim 6, wherein S31 comprises the steps of: S311, initial line head line pixel set aiming at different numbers Wherein Representing the set of coordinates of all pixels of the id numbered line head in the initial image, creating a binary mask If and only if the pixel coordinates Time of day Otherwise, 0; For a pair of Performing morphological dilation operation to obtain Wherein the expanded structuring element is a disc of radius r, Is to A pixel set of any pixel having a Euclidean distance of r or less, and subtracting the line head body region from the expanded region to obtain a surrounding tissue region mask ; S312, extracting the mask of the surrounding tissue area All the pixel coordinates with 1 value form the initial peripheral tissue pixel set Wherein 。
8. The method of orthopedic patient wound care based on visual image processing of claim 7, wherein S32 comprises the steps of: s321, initial pixel number of each line head id Initial pixel number of peripheral tissue Obtaining an initial proportion Obtaining the number of pixels of each line head on the day by performing the same flow as S21-S22 to process the image on the t-th day A wound line pixel set, and the proportion of the thread ends on t days is obtained Wherein A number of pixels of the skin tissue re-identified by color segmentation within the initial surrounding tissue range; Calculating the current rate of change of the ratio Embedded risk early warning threshold based on line head Embedded acknowledgement threshold The thread ends with different numbers are judged every day: When (when) The change of the visibility of the thread ends is not large, and no obvious embedded sign exists; When (when) Then it is indicated that the thread end visibility has begun to drop and the tissue begins to cover the thread end; When (when) Then the representative thread ends are embedded; s322, combining a pixel set of wound lines in the initial image and a t-day wound line pixel set, dividing wound areas under different line areas based on a fixed radius, counting the number of pixels, and calculating the wound shrinkage rate Wherein For an initial image wound pixel count at a fixed radius of the thread end numbered id, Wound pixel count at t days at a fixed radius of the thread end numbered id; s323, aiming at all thread ends under t days, the mark meets the following condition And is also provided with As a detachable tip, wherein Is the wound healing threshold.

Description

Orthopedics patient wound nursing method based on visual image processing Technical Field The invention relates to the technical field of wound care, in particular to a method for nursing wounds of orthopedic patients based on visual image processing. Background In the orthopedic clinical treatment, nursing and stitch removal time after wound suturing are key links for influencing wound healing quality and reducing complication risks, and an orthopedic patient has obvious individual differences due to age, constitution, wound position, injury degree and the like, and the wound healing speed and recovery state show obvious personalized characteristics; The existing method for judging the time of taking out the stitches for wound nursing mainly sets a fixed stitch taking period according to clinical experience, does not fully consider the influence of individual differences of patients on wound healing, is strong in subjectivity, low in accuracy, cannot monitor dynamic changes of wound healing in real time, is difficult to discover abnormality of stitch head areas in time, and has low practicality and functionality because of the problems that the follow-up treatment difficulty is increased, the time of taking out the stitches is slow due to old and weak, basic diseases or wound infection are combined, the healing speed is slow, the early stitch taking can lead to the rupture and bleeding of the wounds, secondary injury and infection risk are caused, and the patients need to frequently go to a hospital to carry out wound review in the existing nursing mode, and medical staff can assist in judging stitch taking time by observing wound states. Disclosure of Invention Aiming at the problems in the related art, the invention provides a method for nursing the wound of an orthopedics patient based on visual image processing, which aims to overcome the technical problems in the prior art. For this purpose, the invention adopts the following specific technical scheme: A method of orthopedic patient wound care based on visual image processing, the method comprising the steps of: S1, constructing a patient information database aiming at an orthopedics patient based on patient basic information, recording the patient basic information and wound information, acquiring daily images of the orthopedics patient after wound suturing through an image acquisition device, and recording the daily images in the patient information database; s2, extracting characteristic pixels of a wound image aiming at the wound image in the patient information database, projecting the characteristic pixels into a blank image, calibrating initial wound line pixels and initial line pixels aiming at the initial image, numbering the line heads, and recording initial line pixels and pixel numbers, initial wound line pixels and pixel numbers; S3, defining a peripheral skin tissue area in an initial line of the initial image, and marking the number of the detachable line head every day according to the number of wound lines of the area of the line head with different numbers on the basis of the number of the pixels of the initial line head and the number of the pixels of the peripheral tissue, and judging the embedded degree of the line head with different numbers by combining the ratio change rates of the line heads with different numbers in the subsequent shooting image. As a preferred embodiment, the step S1 includes the steps of: S11, collecting basic information and wound information of an orthopedic patient, including patient name, contact information, age, wound position, wound length, operation date, stitching date and predicted stitch removal date, establishing a patient information database through MySQL, and establishing a file according to the patient name to record the collected information; S12, acquiring daily images of the wound suture position of the patient through the image acquisition equipment, and uploading the remarking date of the images to the corresponding patient file in the patient information database. As a preferred embodiment, the step S2 includes the steps of: S21, extracting a first patient wound image as an initial image through a current patient file in a patient information database, and identifying a wound and a suture head through an edge detection algorithm; S22, projecting the identified wound lines and the identified line head lines into blank images, recording an initial wound line pixel set and an initial line head line pixel set according to the projected initial images, numbering according to line heads of the current patient wound in sequence, and recording in corresponding patient files. As a preferred embodiment, the step S21 includes the steps of: S211, marking a first patient wound image in a current patient file in a patient information database as an initial image, performing color space conversion, and reducing skin texture noise through Gaussian smoothing filtering to obtain a preprocessed initial i