CN-122005299-A - Human body acupoint automatic identification and projection method, system, equipment and medium

CN122005299ACN 122005299 ACN122005299 ACN 122005299ACN-122005299-A

Abstract

The application relates to a method, a system, equipment and a medium for automatically identifying and projecting human body acupoints, belonging to the technical field of medical image processing, wherein the method comprises the steps of acquiring a real-time patient image and preprocessing; the method comprises the steps of carrying out patient positioning detection on a preprocessed real-time patient image based on a pre-trained target detection model, outputting a positioning posture detection result of a patient, loading an acupoint coordinate template corresponding to the positioning posture detection result when the positioning posture detection result accords with a preset diagnosis and treatment posture, carrying out acupoint recognition on the preprocessed real-time patient image based on the acupoint coordinate template and the pre-trained key point detection model, outputting the current coordinate of a patient acupoint in an image coordinate system, converting the current coordinate into a real position coordinate in a real coordinate system according to a preset projection correction amount, and controlling a laser galvanometer to project a light spot to the patient body based on the real position coordinate so as to mark the acupoint position. The application can improve the accuracy of the acupoint positioning.

Inventors

HE ZHENBANG
YANG PEIHONG
XU YUANKUN
NIU YUN
XU BINGQI

Assignees

陕西汉鑫捷诚科技有限公司

Dates

Publication Date: 20260512
Application Date: 20251125

Claims (9)

1. An automatic human body acupoint recognition and projection method is characterized by comprising the following steps: acquiring a real-time patient image and preprocessing; Performing patient positioning detection on the preprocessed real-time patient image based on a pre-trained target detection model, and outputting a positioning posture detection result of the patient; When the in-position posture detection result accords with a preset diagnosis and treatment posture, loading an acupoint coordinate template corresponding to the in-position posture detection result; Based on the acupoint coordinate template and a pre-trained key point detection model, performing acupoint recognition on the preprocessed real-time patient image, and outputting the current coordinates of the patient acupoint in an image coordinate system; Converting the current coordinate into a real position coordinate in a real coordinate system according to a preset projection correction amount; And controlling the laser galvanometer to project light spots to the body of the patient based on the real position coordinates so as to mark the positions of the acupuncture points.
2. The method for automatically identifying and projecting acupoints on human body according to claim 1, wherein the target detection model comprises CSPDARKNET a backbone network, a feature pyramid network and a decoupling detection head: The step of performing patient in-situ detection on the pre-processed real-time patient image based on a pre-trained target detection model comprises extracting multi-scale features of the real-time patient image through the CSPDARKNET backbone network; fusing the multi-scale features by using the feature pyramid network; Outputting gesture type probability and predicted frame coordinates by the decoupling detection head, wherein the decoupling detection head applies an anchor-free frame technique to compress the number of predicted frames to a single group; And screening a high-confidence prediction frame based on a global optimal matching strategy, and outputting a positioning gesture detection result.
3. The automatic human body acupoint recognition and projection method according to claim 1, wherein the step of performing acupoint recognition on the preprocessed real-time patient image based on the acupoint coordinate template and a pre-trained key point detection model, and outputting the current coordinates of the patient acupoint in the image coordinate system comprises: Inputting the real-time patient image into ResNet main network of the key point detection model to extract image characteristics and outputting a characteristic diagram; Determining a target acupoint region according to the acupoint coordinate template; synchronously generating an initial confidence score and an initial coordinate offset of each acupoint in the target acupoint region on the feature map with the resolution reduced to a preset proportion; According to a preset weighting rule, carrying out fusion calculation on the initial coordinate offset of the current acupoint and the coordinate offsets of all adjacent acupoint to obtain a weighted fusion result; updating the coordinate offset of the current hole site based on the weighted fusion result; and calculating the coordinate positions of each acupoint in the image coordinate system according to the updated coordinate offset, and outputting a coordinate set containing the coordinates of all acupoint as the current coordinate.
4. The automatic human body acupoint recognition and projection method according to claim 1, wherein the projection correction amount is a transformation matrix of an image coordinate system and a real coordinate system, the transformation matrix being defined by a translation parameter (Δx, Δy) and a rotation parameter θ; The projection correction amount satisfies the following equation: Wherein, the The correction amounts are x, y coordinates and an angle θ in the projection correction amounts, and Δx, Δy, and Δθ are x, y coordinate values and an angle value in the translation parameter and the rotation parameter, respectively, x i 、y i 、θ i is an x, y coordinate value and an angle value of a real coordinate point in a real coordinate system, and x d 、y d 、θ d is a coordinate value and an angle value in an image coordinate system corresponding to the real coordinate point.
5. The method for automatically identifying and projecting acupoints on a human body according to any one of claims 1 to 4, further comprising: Acquiring a next frame of real-time patient image; Based on the target detection model, carrying out patient positioning detection on a next frame of real-time patient image, and judging whether a positioning posture detection result changes or not; if yes, terminating the current acupoint recognition flow, updating the acupoint coordinate template according to the changed in-place gesture detection result, and continuously executing the subsequent acupoint recognition steps; If not, carrying out acupoint recognition on the next frame of real-time patient image based on the original acupoint coordinate template and the key point detection model, and updating coordinates of the patient acupoint in an image coordinate system to obtain updated coordinates; calculating the absolute value of the difference between the updated coordinates and the current coordinates; and if the absolute value of the difference is larger than a preset error threshold, converting the updated coordinates into new real position coordinates according to the projection correction amount, and controlling the laser galvanometer to re-project the light spot.
6. The method for automatically identifying and projecting acupoints on a human body according to claim 5, further comprising: When the sitting posture detection result is that no patient or wrong diagnosis and treatment posture exists, generating a posture wrong prompting instruction, and suspending the acupoint recognition and laser projection flow; wait to acquire a new real-time patient image and re-perform the in-situ detection step.
7. An automatic human body acupoint recognition and projection system, comprising: The image acquisition module is used for acquiring a real-time patient image and preprocessing the real-time patient image; The in-position and posture detection module is used for carrying out patient in-position detection on the preprocessed real-time patient image based on a pre-trained target detection model and outputting an in-position and posture detection result of the patient; The acupoint template loading module is used for loading an acupoint coordinate template corresponding to the in-situ posture detection result when the in-situ posture detection result accords with a preset diagnosis and treatment posture; the acupoint coordinate recognition module is used for recognizing the acupoint of the preprocessed real-time patient image based on the acupoint coordinate template and a pre-trained key point detection model, and outputting the current coordinate of the acupoint of the patient in an image coordinate system; the real coordinate conversion module is used for converting the current coordinate into a real position coordinate in a real coordinate system according to a preset projection correction amount; And the laser projection module is used for controlling the laser galvanometer to project light spots to the body of the patient based on the real position coordinates so as to mark the positions of the acupuncture points.
8. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the method of any one of claims 1 to 6 when the program is executed.
9. A computer-readable storage medium, characterized in that a computer program is stored that can be loaded by a processor and that performs the method according to any one of claims 1 to 6.

Description

Human body acupoint automatic identification and projection method, system, equipment and medium Technical Field The application relates to the technical field of medical image processing, in particular to a method, a system, equipment and a medium for automatically identifying and projecting human body acupoints. Background In twenty-first century, the aging process of the population of China is rapidly taking place, the aging condition is more and more serious, and the health care of huge old people becomes a huge social problem, and the aspects of economic competitiveness, social security capability and the like of China are related. The old population generally suffers from certain chronic diseases such as pain, hypertension, diabetes and urinary diseases (such as frequent urination, urgent urination, incomplete urination and the like), and needs to be treated regularly or for a long time, and is healthy, healthful, and the like, and a great deal of medical resources are correspondingly consumed, so that great burden is caused to individuals, families and society. Acupuncture has long been shown to help relieve pain and stress as a traditional chinese medical practice (especially for elderly people suffering from chronic diseases). Therefore, acupuncture is an important and effective alternative medical therapy for sick or disabled elderly people living in areas of low medical coverage. With the deep application of artificial intelligence and automation technology in the medical field, the automatic auxiliary system for traditional Chinese medicine acupuncture treatment becomes a research hotspot, especially in the aspects of acupoint identification and labeling. In the prior art, the automation process of acupuncture treatment faces significant challenges, especially in the acupoint recognition link. Because the human body acupoints are densely distributed and have large individual differences, the acupoints targets are usually smaller in the image and the body surface features are not obvious, the traditional automatic identification method based on the manual design operator has the problems of low robustness and insufficient identification precision. For example, these methods are prone to failure when dealing with changes in the shape or illumination of different patients, resulting in large deviations in the positioning of the acupoints, and cannot meet the requirements for millimeter-level accuracy required in clinic. In addition, because the sitting posture (such as supine, lateral lying or prone lying) of the patient is varied in the acupuncture process, the accuracy of the positioning of the acupuncture points is further reduced. Disclosure of Invention In order to improve the accuracy of acupoint positioning, the application provides a method, a system, equipment and a medium for automatically identifying and projecting human acupoints. In a first aspect, the present application provides a method for automatically identifying and projecting acupoints of a human body, which adopts the following technical scheme: A method for automatically identifying and projecting human body acupoints, the method comprising: acquiring a real-time patient image and preprocessing; Performing patient positioning detection on the preprocessed real-time patient image based on a pre-trained target detection model, and outputting a positioning posture detection result of the patient; When the in-position posture detection result accords with a preset diagnosis and treatment posture, loading an acupoint coordinate template corresponding to the in-position posture detection result; Based on the acupoint coordinate template and a pre-trained key point detection model, performing acupoint recognition on the preprocessed real-time patient image, and outputting the current coordinates of the patient acupoint in an image coordinate system; Converting the current coordinate into a real position coordinate in a real coordinate system according to a preset projection correction amount; And controlling the laser galvanometer to project light spots to the body of the patient based on the real position coordinates so as to mark the positions of the acupuncture points. By adopting the technical scheme, in the embodiment, the image processing, the target detection, the key point identification, the coordinate correction and the laser projection technology are organically combined, so that an automatic and high-precision human body acupoint positioning method is provided. The application can reduce manual intervention in clinical diagnosis and treatment, improve treatment efficiency and ensure treatment accuracy. The system automatically loads the proper acupoint coordinate template and carries out real-time identification and accurate projection by detecting the posture of the patient, avoids the error of the traditional manual marking, has stronger operability and instantaneity, and is particularly suitable for modern medical equ