Medical staff action monitoring and fatigue prevention method and device and storage medium

Abstract

The invention discloses a medical staff action monitoring and fatigue prevention method and device and a storage medium, and relates to the technical field of image processing. The method comprises the following steps: acquiring posture images of medical personnel at different visual angles, and constructing an original data set; performing data enhancement processing on each image in the original data set; marking key points of a human body by utilizing label software; the marked image data set is input into a DNPOSE-YOLOv8 network to be trained; working images of the medical staff are collected, and the trained DNPOSE-YOLOv8 network is used for recognizing coordinate information of all key points of the body of the medical staff during working; calculating the position relation between the key points, and judging whether the working posture of the medical personnel has risks or not; and generating reminding information to indicate the medical staff to optimize the working posture or action. According to the invention, the posture of the medical staff can be rapidly and accurately detected, the bad posture of the medical staff can be corrected, and musculoskeletal diseases can be prevented.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and in particular to a method, device and storage medium for monitoring and preventing the movements of medical personnel. Background Art

[0002] Work-related musculoskeletal disorders (WRMSDs) are a common type of occupational injury caused by prolonged, repetitive or extreme postures, which seriously affect the health and well-being of workers in all walks of life. WRMSDs not only cause health problems for patients, but also cause huge economic losses to individuals and society.

[0003] In order to reduce the hazards caused by WMSDs, researchers conduct ergonomic risk assessments for different work tasks. A common method is to observe videos or pictures of workers at work frame by frame to determine the angles of the upper limbs, lower limbs, neck and torso, as well as factors such as the workers' force and standing posture, and finally complete the risk assessment of the work task. Although this method can help identify and improve factors that may cause WRMSDs in the work environment to a certain extent, its operation process is complicated and time-consuming, and the assessment results are greatly affected by subjective factors, which limits the assessment accuracy and intervention effect. Summary of the invention

[0004] In view of the problem that the existing technology is inaccurate and untimely in monitoring the posture of medical staff, the present invention provides a method, device and storage medium for monitoring the movement and preventing fatigue of medical staff, so as to realize rapid and accurate detection of the posture of medical staff, correct the bad posture of medical staff and prevent musculoskeletal diseases.

[0005] In order to achieve the above object, the present invention adopts the following technical solution: According to one aspect of the present invention, a method for monitoring the movements and preventing fatigue of medical staff is provided, comprising: acquiring posture images of medical staff from different perspectives acquired by a high-precision camera to construct an original data set; performing data enhancement processing on each image in the original data set to increase data diversity and expand the data set; marking key points of the human body on each image using labelme software; inputting the labeled image data set into a DNPOSE-YOLOv8 network for training to obtain a trained DNPOSE-YOLOv8 network, wherein the trained DNPOSE-YOLOv8 network is used to identify a target object in an input image and locate the positions of key points of the target object; acquiring working images of medical staff in real time, and using the trained DNPOSE-YOLOv8 network to identify coordinate information of key points of the body of the medical staff when working; calculating the positional relationship between the key points based on the coordinate information, and judging whether there is a risk in the working posture of the medical staff based on the positional relationship between the key points; generating a reminder message when a risky posture is detected to instruct the medical staff to optimize the working posture or movement to reduce occupational fatigue.

[0006] Optionally, data enhancement processing is performed on each image in the original data set, including: rotating each image by ±30°, ±45°, ±60°, and ±90° in a clockwise direction; performing random brightness and darkness adjustment on each image, and screening the images after the brightness and darkness changes to remove images whose brightness does not meet the requirements; and adding salt and pepper noise, Gaussian noise, and speckle noise to each image.

[0007] Optionally, before inputting the annotated image dataset into the DNPOSE-YOLOv8 network for training, the method further includes: uniformly adjusting the image size in the dataset to 640*640, and scaling the pixel value to between 0 and 1.

[0008] Optionally, the DNPOSE-YOLOv8 network is an improvement based on YOLOv8-pose, and the improvements include: adopting a lighter detection head LW-Pose to speed up the network's reasoning process; introducing an asymmetric design in the network structure to reduce computational complexity and the amount of model parameters; using a 3×3 depthwise separable convolution DS_Conv instead of a traditional 3×3 convolution in each branch to reduce the number of parameters; adding a feature-focused diffusion pyramid structure to the HEAD network structure, the pyramid structure extracts features from each layer and performs feature fusion, and then diffuses the fused features to each feature dimension, so that each feature dimension has more comprehensive feature information, thereby realizing the feature-focused diffusion of the pyramid structure and improving detection accuracy.

[0009] Optionally, the pyramid structure adopts a feature focusing module and a feature diffusion mechanism, wherein the feature focusing module is intended to capture and integrate feature information from different scales and generate a feature map with contextual information, wherein the feature focusing module accepts inputs of three scales and internally includes an Inception-Style module, and the Inception-Style module uses a set of parallel deep convolutions to capture information across multiple scales; the feature diffusion mechanism is intended to diffuse the contextual information features generated by the feature focusing module to each detection scale, so as to achieve cross-scale feature fusion and enhancement and improve detection accuracy.

[0010] Optionally, the positional relationship between the key points includes the distance between the key points, and calculating the positional relationship between the key points according to the coordinate information, and judging whether there is a risk in the working posture of the medical staff according to the positional relationship between the key points includes: obtaining the coordinates of the left and right ankles output by the DNPOSE-YOLOv8 network, and calculating the normal distance between the left and right ankles; judging whether the normal distance between the left and right ankles exceeds a preset threshold, and if so, it is considered that there is a risk in the working posture of the medical staff.

[0011] Optionally, the positional relationship between the key points includes joint angles, and the calculating the positional relationship between the key points according to the coordinate information, and judging whether there is a risk in the working posture of the medical staff according to the positional relationship between the key points includes: For the three key points that make up the joint angle , , ,in, For the joint point, , For joint points Adjacent key points, line segments and The angle formed is the joint The joint angles are based on , , Coordinate value calculation vector , : ; ; In the formula, Key points The x, y, and z coordinate values ​​of Key points The x, y, and z coordinate values ​​of Key points The x, y, and z coordinate values ​​of Calculate vectors by vector dot product and modulus and The angle between ω The cosine of : ; make , then the joint Joint angle ; It is determined whether the calculated joint angle is within the preset range. If not, it is considered that the medical staff's working posture is risky.

[0012] Optionally, generating reminder information when a risk posture is detected includes: setting risk levels, including negligible risk, low risk, medium risk and high risk, wherein negligible risk corresponds to RULA level 1-2, low risk corresponds to RULA level 3-4, medium risk corresponds to RULA level 5-6, and high risk corresponds to RULA level 7; displaying reminder information corresponding to the risk level in fonts of different colors through a display interface, wherein negligible risk corresponds to white font, low risk corresponds to yellow font, medium risk corresponds to orange font, and high risk corresponds to red font.

[0013] According to another aspect of the present invention, a device for monitoring the movements of medical staff and preventing fatigue is also provided. The device comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, the steps of the method described above are implemented.

[0014] According to another aspect of the present invention, a storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method described above are implemented.

[0015] The present invention proposes a method for monitoring the movements and preventing fatigue of medical staff, which realizes the rapid and accurate detection of the posture of medical staff, avoids human subjective errors, can correct the bad posture of medical staff, and prevent musculoskeletal diseases. In addition, the method of the present invention avoids the physical and psychological discomfort caused to workers by wearable sensors when working, has higher automation and more detailed and objective evaluation, and reduces the time-consuming and complicated disadvantages of the evaluation method. The present invention has the characteristics of high efficiency, real-time and accuracy, and can be widely used in medical scenarios to help medical staff optimize the posture and movements of dispensing medicine, reduce occupational fatigue, and improve work efficiency and safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings: Figure 1 is a flow chart of a method for monitoring movement and preventing fatigue of medical staff according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the camera layout; Figure 3 This is the DNPOSE-YOLOv8 network structure diagram; Figure 4 This is the LW-Pose structure diagram; Figure 5 This is a feature focused diffusion pyramid structure diagram; Figure 6 This is the structural diagram of the feature focusing diffusion module; Figure 7 The following are the experimental results of the original YOLOv8-pose model and the DNPOSE-YOLOv8 model. DETAILED DESCRIPTION

[0017] In order to enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only embodiments of a part of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work should fall within the scope of protection of the present application. It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict.

[0018] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0019] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0020] Embodiment 1: This embodiment provides a method for monitoring the movements and preventing fatigue of medical staff, which is intended to detect whether there are problems with the movements and postures of medical staff during work, so as to reduce physical fatigue and overwork caused by bad movements.

[0021] Reference Figure 1 , Figure 1 FIG. 1 is a flow chart of a method for monitoring the movements of medical personnel and preventing fatigue according to an embodiment of the present invention. Figure 1 As shown, the method comprises the following steps: S1, obtain the posture images of medical staff from different perspectives collected by high-precision cameras and construct the original data set; High-precision camera layout can be referenced Figure 2 , ensure that the camera's shooting range should cover the working area of ​​medical staff, so as to be able to fully and accurately capture the various movements and postures of medical staff during their work.

[0022] S2, perform data augmentation on each image in the original dataset to increase data diversity and expand the dataset; Data enhancement processing includes rotation, brightness and darkness adjustment, and noise addition, etc., which aims to simulate various possible shooting conditions and environmental changes, such as lighting, angle, noise, etc., so that the model can better adapt to various situations in practical applications.

[0023] In the embodiment of the present invention, the data enhancement processing includes: S21, taking the clockwise direction as the positive direction, rotating each image by ±30°, ±45°, ±60°, and ±90° respectively; By rotating the image, we can simulate the posture changes under different viewing angles, which helps the model learn the human posture characteristics under different angles. In this embodiment, each original image is rotated into 8 different angles (including the original unrotated version, a total of 9), and the rotated images will be used as new training samples to increase the diversity of the data set.

[0024] S22, performing random brightness and darkness adjustment on each image, and screening the images after the brightness and darkness change to remove images whose brightness does not meet the requirement; By adjusting the brightness and darkness, we can simulate shooting scenes under different lighting conditions, which helps the model adapt to images under different light intensities and improves its robustness in real environments. Random brightness and darkness adjustment can be achieved by randomly adjusting the brightness and contrast of the image.

[0025] In addition, in order to avoid the loss of image information caused by overexposure or too dark, the embodiment of the present invention also screens the images after the brightness change, and removes the overexposed or too dark images by checking the image contrast, brightness and other indicators, or whether the key information in the image is still visible, to ensure that the image quality in the data set meets the training requirements.

[0026] S23, adding salt and pepper noise, Gaussian noise, and speckle noise to each image.

[0027] Adding noise is to simulate the interference that the image may be subject to during transmission or acquisition. By adding noise, the model will learn how to accurately identify human posture in the presence of noise, thereby improving its detection ability in low-quality images and improving the robustness of the model. Among them, salt and pepper noise, also known as impulse noise, randomly adds black and white noise to the image, sets some pixels to black or white, and simulates random bright or dark spots in the image; Gaussian noise randomly changes the value of pixels in the image according to Gaussian distribution (normal distribution) to simulate random fluctuations in the image; speckle noise simulates speckle-like interference in the image and simulates texture changes in the image.

[0028] The above data enhancement steps can increase the diversity and complexity of the data set, help the model learn more generalized feature representations, adapt to various changing conditions, and thus improve its stability and reliability in real application scenarios. For example, in one specific implementation, after the above data enhancement process, the original data set is expanded to 3,000 images.

[0029] S3, label the key points of the human body in each image using labelme software; The key points of the human body are the main joints of the human body, such as the head, shoulders, elbows, wrists, hips, knees and ankles. Other key points may also be included according to specific application requirements, such as fingertips, toe tips or facial feature points. By marking the key points of the human body, the model can accurately identify and understand the position and relationship of various parts of the human body in space.

[0030] In one embodiment, the number of key points of the human body is set to 13, namely, head (1), neck (1), shoulder (2 left and right), elbow (2 left and right), wrist (2 left and right), hip (1), knee (2 left and right) and ankle (2 left and right). The above 13 key points of the human body are annotated for each image in the dataset through labelme software. These images with annotated key points will be used as model training samples to train the DNPOSE-YOLOv8 network model. The model will learn how to identify and locate the key points of the human body based on the input image.

[0031] Furthermore, before sending the image to the DNPOSE-YOLOv8 network for training, the image needs to be preprocessed, including image size unification and pixel normalization. The image size is uniformly adjusted to 640*640 pixels by using image scaling algorithms (such as bilinear interpolation, nearest neighbor interpolation, etc.) to meet the input requirements of the DNPOSE-YOLOv8 network; and the image pixel values ​​are scaled to between 0 and 1, and the RGB color value of each pixel (usually an integer ranging from 0-255) is converted, and each pixel value is divided by 255 to become a floating point number between 0 and 1. After the image preprocessing steps, each image in the dataset will have a uniform size (640x640) and pixel value range (0 to 1). Such a dataset is more suitable for training the DNPOSE-YOLOv8 network, and the network can more easily learn and extract features in the image.

[0032] The dataset after size unification and pixel normalization will be divided into training set, validation set and test set. The training set is used to train the network model, the validation set is used to evaluate the performance of the model and adjust the hyperparameters during the training process, and the test set is used to evaluate the final performance of the model after the training is completed.

[0033] S4, inputting the annotated image data set into the DNPOSE-YOLOv8 network for training to obtain a trained DNPOSE-YOLOv8 network, wherein the trained DNPOSE-YOLOv8 network is used to identify the target object in the input image and locate the position of each key point of the target object; DNPOSE-YOLOv8 is an advanced network model that integrates human posture detection and target detection functions. It can not only efficiently identify the target objects (i.e. medical staff) in the image, but also accurately locate the positions of various key points of the human body.

[0034] Reference Figure 3 , Figure 3 DNPOSE-YOLOv8 network structure diagram. The DNPOSE-YOLOv8 network in the embodiment of the present invention is an improvement based on YOLOv8-pose, including: 1. Use a lighter detection head LW-Pose ( Figure 4 LW-Pose structure diagram) to speed up the network reasoning process; 2. Introduce asymmetric design in the network structure to reduce computational complexity and model parameter quantity; Asymmetric design is a deep learning network structure optimization method, which refers to introducing asymmetry in the network's convolution kernel, pooling window, or connection mode. By using convolution kernels of different sizes, asymmetric pooling windows, or adjusting the connection mode of the network layer, such as using non-square convolution kernels 1*3 or 3*5, instead of using common symmetric designs (such as standard 3*3 or 5*5 convolution kernels), the amount of calculation can be reduced, the reasoning speed and feature expression ability can be improved, and computing resources can be used more efficiently.

[0035] 3. Use 3×3 depth-separable convolution DS_Conv in each branch instead of traditional 3×3 convolution to reduce the number of parameters; DS_Conv decomposes the convolution operation into Depthwise Convolution and pointwise convolution. The number of feature maps after Depthwise Convolution is the same as the number of channels in the input layer, and the feature map cannot be expanded. Since this operation is performed independently on each channel of the input layer, it is impossible to effectively utilize the feature information of different channels at the same spatial position. Therefore, pointwise convolution is needed to merge these feature maps to generate a new feature map. However, since it cannot effectively utilize the feature information of different channels at the same spatial position, pointwise convolution will be used for dimension expansion later. The operation of pointwise convolution is very similar to the conventional convolution operation. It mixes the information between channels by applying a 1×1 convolution kernel between different channels. It does not change the spatial dimension of the feature map, but only changes the number of channels. Therefore, the pointwise convolution operation weights and merges the previous images in the depth direction to generate a new feature map. Secondly, due to the reduction in the number of parameters, the depthwise separable convolution is more efficient in calculation and suitable for use in resource-limited environments. Finally, the information of different channels is combined through pointwise convolution to maintain a certain feature extraction capability.

[0036] For example, the convolution kernel size is D K * D K , the number of input channels isM , the number of output channels is N , the output feature map size is D F * D F , after standard convolution, we can calculate: The parameter quantities are: ; The amount of calculation is: ; Using DS_Conv of the same size in this embodiment, it can be calculated: The parameter quantities are: ; The amount of calculation is: ; By comparison, it can be seen that the depthwise separable convolution adopted in the embodiment of the present invention has significantly fewer parameters and calculation amount than the standard convolution. Therefore, the model parameters can be effectively reduced and the detection speed can be improved.

[0037] 4. Add feature-focused diffusion pyramid structure to the HEAD network structure ( Figure 5 ), the pyramid structure extracts the features of each layer and performs feature fusion, and then diffuses the fused features to each feature dimension, so that each feature dimension has more comprehensive feature information, thereby realizing the feature focus diffusion of the pyramid structure and improving the detection accuracy.

[0038] The feature focus diffusion pyramid structure adopts a customized feature focus module and feature diffusion mechanism, where: Figure 6 This is a structural diagram of the feature focus diffusion module. The feature focus module is designed to capture and integrate feature information from different scales and generate a feature map with contextual information. The feature focus module accepts inputs at three scales, i.e., detects targets of different sizes at different feature map scales. It contains an Inception-Style module, which uses a set of parallel deep convolutions to capture information across multiple scales. The feature diffusion mechanism is designed to diffuse the contextual information features generated by the feature focus module to each detection scale, so as to achieve cross-scale feature fusion and enhancement and improve detection accuracy.

[0039] The original YOLOv8-pose model and the DNPOSE-YOLOv8 model proposed in the embodiment of the present invention are used for ablation experiment comparison. The experimental results are as follows: Figure 7 As shown, Figure 7Pose-mAP@0.5 is a measure of the mAP (average precision) of key point detection under the threshold of IoU (intersection over union) = 0.5, and Pose-mAP@0.95 is a measure of the mAP (average precision) of key point detection under the threshold of IoU = 0.95. FLOPs / G represents the complexity of the network, and Params / M is the number of network parameters in M. Compared with the original YOLOv8-pose model, the DNPOSE-YOLOv8 model proposed in the embodiment of the present invention has advantages in various indicators, which verifies that the improvement is effective.

[0040] S5, real-time collection of working images of medical staff, and use of the trained DNPOSE-YOLOv8 network to identify the coordinate information of key points of the medical staff's body when they are working The DNPOSE-YOLOv8 network generates key points for posture detection through the detection head. The feature map output by the network is converted into the coordinates and size of the bounding box, the key points for posture detection, and the probability density distribution is converted into the category probability to obtain the final bounding box coordinates, category probability, confidence score and key point coordinate information. After the image is output by the network, each bounding box and the key point coordinates contained in the box are saved to the specified folder through image cropping technology. Each file contains a medical staff working posture image and the corresponding key point coordinates.

[0041] S6, calculating the positional relationship between the key points according to the coordinate information, and judging whether there is a risk in the working posture of the medical staff according to the positional relationship between the key points; Traverse the images in the specified folder (i.e. the folder where the DNPOSE-YOLOv8 network output results are located), read each image and its corresponding key point coordinate information, calculate the positional relationship between the key points based on the coordinate information of the key points output by the model, the positional relationship includes distance and angle, and judge whether there is a risk in the working posture of the medical staff based on the calculated positional relationship.

[0042] In one embodiment, S6 includes risk assessment based on the normal distance between the left and right ankles, specifically: S611, obtaining the coordinates of the left and right ankles output by the DNPOSE-YOLOv8 network, and calculating the normal distance between the left and right ankles; S612, determining whether the normal distance between the left and right ankles exceeds a preset threshold. If so, it is considered that the working posture of the medical staff is risky.

[0043] Normal distance refers to the distance between the left and right ankles in the direction perpendicular to the ground. This distance can be used to assess the standing stability of medical staff or whether there is a risk of falling. The standing situation of the medical staff is monitored in real time and judged based on the spatial position information. When the difference in the normal distance between the two ankles is greater than the preset threshold (such as 20mm), it is judged as standing on one foot and there is a risk.

[0044] In another embodiment, S6 includes risk assessment based on joint angles, specifically: The vector method is used to calculate the joint angles of the key points in each image. The vector has directionality and translation, so any two non-overlapping points in the human body coordinate system , Can be converted into a vector in the reference coordinate system in the form of a coordinate vector through coordinate transformation .

[0045] S621, for the three key points that make up the joint angle , , ,in, For the joint point, , For joint points Adjacent key points, line segments and The angle formed is the joint The joint angles are based on , , Coordinate value calculation vector , : ; ; S622, Calculate vectors by vector dot product and modulus , The angle between ω The cosine of : ; make , then the joint Joint angle ; S623, determining whether the calculated joint angle is within a preset range. If not, it is considered that the medical staff's working posture is risky.

[0046] Based on medical or ergonomic knowledge, a preset range is set for each joint angle. For example, the normal flexion of the cervical spine is generally 35 to 45 degrees. If the calculated joint angle is not within the preset range, it is considered that the medical staff's working posture is risky, such as excessive extension or bending of the joint.

[0047] It can be understood that in addition to the above-mentioned ankle distance and joint angle, in other embodiments, the positional relationship between the left and right knees can also be calculated, such as whether they overlap, cross, etc., to determine whether there is an unhealthy posture such as crossing the legs, or to calculate the bending amplitude of the cervical spine and waist to determine whether there is a hunchback, etc.

[0048] S7, when a risky posture is detected, generates a reminder message to instruct medical staff to optimize their working posture or movements to reduce occupational fatigue.

[0049] Reminder information includes text, sound or images, and can be conveyed to medical staff through display screens, speakers or mobile devices.

[0050] In one embodiment, risk levels are set, including negligible risk, low risk, medium risk and high risk, wherein negligible risk corresponds to RULA (Rapid Upper Limb Assessment) level 1-2, low risk corresponds to RULA level 3-4, medium risk corresponds to RULA level 5-6, and high risk corresponds to RULA level 7; different risk ranges are defined, and when the calculated positional relationship is at different risk levels, reminder information corresponding to the risk level is displayed in different color fonts through the display interface, wherein negligible risk corresponds to white font, low risk corresponds to yellow font, medium risk corresponds to orange font, and high risk corresponds to red font. By distinguishing different colors, medical staff can be warned more intuitively.

[0051] Through the above steps, the present invention can efficiently, real-time and accurately complete the posture assessment of medical staff, and remind risky postures. The present invention has the advantages of high automation, detail and objectivity, and can be widely used in medical scenarios to help medical staff optimize working postures and movements (such as dispensing movements), reduce occupational fatigue, and improve work efficiency and safety.

[0052] Example 2: With respect to the same inventive concept, a medical staff motion monitoring and fatigue prevention device is also provided in this example, which is used to implement the above examples and preferred implementation methods, and those that have been explained will not be repeated here.

[0053] For specific examples in this embodiment, reference may be made to the examples described in the above embodiments and optional implementation modes, and this embodiment will not be described in detail here.

[0054] An embodiment of the present invention further provides a storage medium, in which a computer program is stored, wherein the computer program is configured to execute the steps of any of the above method embodiments when running.

[0055] Optionally, in this embodiment, the above-mentioned storage medium may include but is not limited to: a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk, and other media that can store computer programs.

[0056] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0057] In the above embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0058] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for monitoring the movements of medical staff and preventing fatigue, characterized in that: include: Obtain posture images of medical staff from different perspectives captured by high-precision cameras and construct the original data set; Perform data augmentation on each image in the original dataset to increase data diversity and expand the dataset; Use labelme software to mark the key points of the human body in each image; Input the labeled image data set into the DNPOSE-YOLOv8 network for training to obtain a trained DNPOSE-YOLOv8 network, wherein the trained DNPOSE-YOLOv8 network is used to identify the target object in the input image and locate the position of each key point of the target object; Collecting working images of medical staff in real time, and using the trained DNPOSE-YOLOv8 network to identify coordinate information of key points of the medical staff's body when they are working; Calculate the positional relationship between the key points according to the coordinate information, and determine whether there is a risk in the working posture of the medical staff according to the positional relationship between the key points; When a risky posture is detected, a reminder message is generated to instruct medical staff to optimize their working posture or movements to reduce occupational fatigue.

2. The method for monitoring the movement and preventing fatigue of medical staff according to claim 1, characterized in that: The data augmentation process for each image in the original dataset includes: Taking clockwise as the positive direction, each image is rotated by ±30°, ±45°, ±60°, and ±90° respectively; Perform random brightness and darkness adjustment on each image, and screen the images after brightness and darkness change to remove images whose brightness does not meet the requirements; Salt and pepper noise, Gaussian noise, and speckle noise are added to each image.

3. The method for monitoring the movement and preventing fatigue of medical personnel according to claim 1, characterized in that: Before inputting the labeled image dataset into the DNPOSE-YOLOv8 network for training, the method further includes: The images in the dataset are uniformly resized to 640*640 and the pixel values ​​are scaled to between 0 and 1.

4. The method for monitoring the movement and preventing fatigue of medical personnel according to claim 1, characterized in that: The DNPOSE-YOLOv8 network is an improvement on the YOLOv8-pose. The improvements include: Use a lighter detection head, LW-Pose, to speed up the network’s reasoning process; Introducing asymmetric design into the network structure to reduce computational complexity and model parameter count; In each branch, 3×3 depth-wise separable convolution DS_Conv is used instead of traditional 3×3 convolution to reduce the number of parameters; A feature focused diffusion pyramid structure is added to the HEAD network structure. The pyramid structure extracts features from each layer and performs feature fusion, and then diffuses the fused features to each feature dimension so that each feature dimension has more comprehensive feature information, thereby realizing feature focused diffusion of the pyramid structure and improving detection accuracy.

5. The method according to claim 4, characterized in that The pyramid structure adopts a feature focusing module and a feature diffusion mechanism. The feature focusing module is designed to capture and integrate feature information from different scales and generate a feature map with contextual information. The feature focusing module accepts inputs of three scales and contains an Inception-Style module. The Inception-Style module uses a set of parallel deep convolutions to capture information across multiple scales. The feature diffusion mechanism is designed to diffuse the contextual information features generated by the feature focusing module to each detection scale to achieve cross-scale feature fusion and enhancement and improve detection accuracy.

6. The method according to claim 1, characterized in that The positional relationship between the key points includes the distance between the key points, and the calculating the positional relationship between the key points according to the coordinate information, and judging whether there is a risk in the working posture of the medical staff according to the positional relationship between the key points includes: Get the coordinates of the left and right ankles output by the DNPOSE-YOLOv8 network and calculate the normal distance between the left and right ankles; Determine whether the normal distance between the left and right ankles exceeds the preset threshold. If so, it is considered that the medical staff's working posture is risky.

7. The method according to claim 1, characterized in that The positional relationship between the key points includes the joint angles, and the calculating the positional relationship between the key points according to the coordinate information, and judging whether there is a risk in the working posture of the medical staff according to the positional relationship between the key points includes: For the three key points that make up the joint angle , , ,in, For the joint point, , For joint points Adjacent key points, line segments and The angle formed is the joint The joint angles are based on , , Coordinate value calculation vector , : ； ； In the formula, Key points The x, y, and z coordinate values ​​of Key points The x, y, and z coordinate values ​​of Key points The x, y, and z coordinate values ​​of Calculate vectors by vector dot product and modulus and The angle between ω The cosine of : ； make , then the joint Joint angle ; It is determined whether the calculated joint angle is within the preset range. If not, it is considered that the medical staff's working posture is risky.

8. The method according to claim 1, characterized in that The generating of reminder information when a risky posture is detected includes: Set risk levels, including negligible risk, low risk, medium risk and high risk, where negligible risk corresponds to RULA level 1-2, low risk corresponds to RULA level 3-4, medium risk corresponds to RULA level 5-6, and high risk corresponds to RULA level 7; The display interface displays the corresponding risk level reminder information in different color fonts, where negligible risks are displayed in white fonts, low risks are displayed in yellow fonts, medium risks are displayed in orange fonts, and high risks are displayed in red fonts.

9. A medical staff motion monitoring and fatigue prevention device, characterized in that: The medical staff motion monitoring and fatigue prevention device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, the steps of the method described in any one of claims 1 to 8 are implemented.

10. A storage medium, characterized in that: The storage medium stores a computer program, which implements the steps of the method according to any one of claims 1 to 8 when executed by a processor.

Images