Intelligent image auxiliary guidance system for lung cancer postoperative care

Abstract

The invention relates to the technical field of image recognition, in particular to a lung cancer postoperative care-oriented intelligent image auxiliary guidance system, which is characterized in that real change information values of lung clusters are determined by acquiring lung CT images of a lung cancer patient at each sampling moment after operation, so that suspected infected lung clusters are screened out, and the lung cancer postoperative care-oriented intelligent image auxiliary guidance system is obtained. And determining regularity of a lung cluster region boundary, adjusting the actual change information value according to the regularity, and determining an actual change result of a suspected infected lung cluster so as to determine the infection rate of the lung cancer patient at each sampling moment, thereby obtaining diffusion attributes of the lung cancer patient and giving different levels of nursing guidance suggestions to the lung cancer patient. By analyzing the lung CT images of the patient in different periods, the corresponding infection prediction diffusion results and the corresponding nursing guidance suggestions are obtained, the risk of postoperative complications is reduced, and the nursing accuracy is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of image recognition, and in particular to an intelligent image-assisted guidance system for postoperative nursing of lung cancer. Background Art

[0002] Postoperative nursing guidance for lung cancer can monitor the patient's recovery and provide efficient recovery care plans. Among them, postoperative care mainly includes: postoperative recovery monitoring, image-assisted identification monitoring, lung function recovery assessment, and complication prevention and care. After lung cancer surgery, the patient's recovery process needs to be closely monitored, including the healing of the surgical site and the recovery of lung function. Imaging examinations, such as chest CT, X-rays, and lung function tests, are important means of monitoring the patient's postoperative recovery, and the specific nursing goal is to monitor postoperative complications such as lung infection, pneumothorax, and mediastinal effusion through the analysis of imaging data. The intelligent imaging system can detect postoperative tissue changes, detect abnormalities in a timely manner, and help nurses adjust the nursing plan in a timely manner.

[0003] Existing problems: Postoperative care for lung cancer is particularly important, because after surgery, the patient's immune system is suppressed in many ways, which provides a breeding ground for lung infections. The specific reasons for the suppression are: intraoperative chemotherapy and the immunosuppressive effects of radiotherapy. Lung cancer patients often receive chemotherapy. Chemotherapy drugs will suppress the function of the bone marrow, resulting in a decrease in the number of white blood cells (especially neutrophils), which in turn affects the immune response, greatly reducing the body's ability to resist infections such as bacteria and viruses, and making lung infections more likely to occur. In addition, in order to prevent postoperative complications, some patients will use immunosuppressive drugs (such as glucocorticoids, etc.) for a long time. Such drugs will further suppress the immune system, making patients more susceptible to infection. Therefore, when the analysis of the patient's lung infection based on lung CT images is inaccurate, it will lead to inappropriate auxiliary nursing guidance suggestions for patients, which may reduce the patient's recovery effect after lung cancer surgery. Summary of the invention

[0004] The present invention provides an intelligent image-assisted guidance system for postoperative nursing of lung cancer to solve the existing problems.

[0005] The intelligent image-assisted guidance system for postoperative nursing of lung cancer of the present invention adopts the following technical solutions:

[0006] An embodiment of the present invention provides an intelligent image-assisted guidance system for postoperative nursing of lung cancer, the system comprising the following modules:

[0007] Lung CT image acquisition module: used to obtain lung CT images of lung cancer patients at each sampling moment after surgery;

[0008] Lung consolidation information analysis module: used to divide the lung CT images into several lung clusters, and determine the consolidation information value of the lung cluster according to the number of pixels and the size of the gray value in the lung cluster;

[0009] Lung consolidation result determination module: used to screen out a number of suspected infected lung clusters according to the size of the consolidation information value; record the connected area formed by adjacent pixels in each suspected infected lung cluster as a lung cluster area; determine the regularity of the lung cluster area boundary according to the distribution of pixels on the lung cluster area boundary; adjust the consolidation information value of the suspected infected lung cluster according to the regularity, and determine the consolidation result of the suspected infected lung cluster;

[0010] Nursing guidance recommendation determination module: used to determine the infection rate of lung cancer patients at each sampling time according to the size of the solid change result; determine the diffusion property of lung cancer patients according to the difference in infection rate of lung cancer patients at adjacent sampling times; and give different levels of nursing guidance recommendations to lung cancer patients according to the size of the diffusion property.

[0011] Further, determining the consolidation information value of the lung cluster includes:

[0012] In any one lung cluster, the mean of the grayscale values ​​of all pixels is obtained, the inversely proportional normalized value of the sum of the absolute values ​​of the differences between the grayscale values ​​of all pixels and the mean is obtained, the product of the inversely proportional normalized value and the mean is calculated, and the product and the normalized value of the sum of the number of pixels in the any one lung cluster are used as the real change information value of the any one lung cluster.

[0013] Furthermore, the screening of several suspected lung infection clusters includes:

[0014] When the consolidation information value of any lung cluster is greater than a preset infection threshold, the any lung cluster is recorded as a suspected infected lung cluster.

[0015] Further, the determining of the regularity of the lung cluster region boundary includes:

[0016] For any lung cluster region, obtain the mean of the Euclidean distances between the central pixel of the lung cluster region and all the pixel points on the boundary of the lung cluster region, and construct a circular region with the central pixel of the lung cluster region as the dot and the mean as the radius;

[0017] The regularity of the lung cluster region boundary is determined based on the overlap between the lung cluster region boundary and the circular region boundary and the Euclidean distance from the central pixel point of the lung cluster region to the lung cluster region boundary.

[0018] Further, the determining the regularity of the lung cluster region boundary according to the overlap between the lung cluster region boundary and the circular region boundary and the Euclidean distance from the central pixel point of the lung cluster region to the lung cluster region boundary includes:

[0019] The number of overlapping pixels between the boundary of the lung cluster region and the boundary of the circular region is obtained, and an inversely proportional normalized value of the sum of the absolute values ​​of the differences between the Euclidean distances between the central pixel of the lung cluster region and all the pixels on the boundary of the lung cluster region and the mean is calculated; and a normalized value of the product of the inversely proportional normalized value and the number of overlapping pixels is used as the regularity of the boundary of the lung cluster region.

[0020] Further, the determination of the consolidation results of the suspected infected lung clusters includes:

[0021] Taking the average of the regularities of all lung cluster region boundaries within any suspected infected lung cluster as the regularity of the suspected infected lung cluster;

[0022] The normalized value of the ratio of the consolidation information value to the regularity of the suspected infected lung cluster is used as the consolidation result of the suspected infected lung cluster.

[0023] Further, determining the infection rate of lung cancer patients at each sampling time includes:

[0024] The mean of the consolidation results of all suspected infected lung clusters in the lung CT images at any sampling time is taken as the infection rate of lung cancer patients at the sampling time.

[0025] Further, determining the diffusion attribute of the lung cancer patient includes:

[0026] Get the The sampling time and The difference in infection rates of lung cancer patients at different sampling times ,like , then assign the Sampling time to The label of the sampling moment is 0. , then assign the Sampling time to The label of each sampling moment is 1. , then assign the Sampling time to The label of the sampling moment is -1;

[0027] Determine the diffusion properties of lung cancer patients based on the labels of all adjacent sampling moments.

[0028] Furthermore, determining the diffusion attribute of the lung cancer patient according to the labels of all adjacent sampling moments includes:

[0029] The normalized value of the sum of the labels at all adjacent sampling moments is recorded as the diffusion attribute of the lung cancer patient.

[0030] Furthermore, the providing of different levels of nursing guidance suggestions to lung cancer patients according to the magnitude of the diffusion attribute includes:

[0031] When the diffusion attribute of a lung cancer patient is greater than a preset diffusion threshold, the lung cancer patient is given first-level nursing guidance suggestions; when the diffusion attribute of a lung cancer patient is less than or equal to the preset diffusion threshold, the lung cancer patient is given second-level nursing guidance suggestions.

[0032] The beneficial effects of the technical solution of the present invention are:

[0033] In an embodiment of the present invention, a lung CT image of a lung cancer patient at each sampling moment after surgery is obtained, and a number of lung clusters are divided, and the consolidation information value of the lung cluster is determined to screen out a number of suspected infected lung clusters, and a connected area composed of adjacent pixels in each suspected infected lung cluster is recorded as a lung cluster area, and the regularity of the lung cluster area boundary is determined. The consolidation information value is adjusted according to the regularity to determine the consolidation result of the suspected infected lung cluster to determine the infection rate of the lung cancer patient at each sampling moment, thereby analyzing the CT images of the patient at different periods, combined with the lung infection characteristics of different regions in the image, such as the local high density phenomenon caused by consolidation, to obtain the infection rate of the lungs, and judge the specific lesion attributes of the patient based on the infection rate, thereby improving the authenticity of the analysis and identification results. The diffusion properties of lung cancer patients are obtained to provide different levels of nursing guidance and suggestions for lung cancer patients. By analyzing the lung infection results of patients at different stages, the corresponding lung infection diffusion properties of patients at multiple stages are obtained, and the diffusion properties are used as the factors of auxiliary nursing guidance opinions, which plays the nursing purpose of preventing complications from occurring in postoperative nursing of patients, and improves the nursing accuracy and nursing efficiency. So far, the present invention obtains the corresponding infection prediction diffusion results and corresponding nursing guidance and suggestions by analyzing the lung CT images of patients at different stages, reduces the risk of postoperative complications, and improves the accuracy of nursing. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0035] Figure 1This is a module flow chart of an intelligent image-assisted guidance system for postoperative nursing of lung cancer according to the present invention;

[0036] Figure 2 Obtaining flow charts for nursing guidance recommendations;

[0037] Figure 3 Schematic diagram of the lung cluster area and the corresponding circular area. DETAILED DESCRIPTION

[0038] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation method, structure, features and effects of an intelligent image-assisted guidance system for postoperative care of lung cancer proposed by the present invention are described in detail below in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0039] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0040] The specific scheme of an intelligent image-assisted guidance system for postoperative nursing of lung cancer provided by the present invention is described in detail below with reference to the accompanying drawings.

[0041] See also Figure 1 , which shows a module flow chart of an intelligent image-assisted guidance system for lung cancer postoperative care provided by an embodiment of the present invention, the system includes the following modules:

[0042] Module 101: Lung CT image acquisition module.

[0043] This module is used to obtain lung CT images of lung cancer patients at each sampling moment after surgery.

[0044] Collect lung CT images of any lung cancer patient at each sampling time after surgery.

[0045] It should be noted that the lung CT image is a grayscale image, and the sampling time is selected as twice a month after surgery for lung cancer patients until the current moment when the patient has not recovered. This example is used for description. Thus, through multiple lung CT image acquisitions, the dynamic changes of the lungs can be captured, which can be used to evaluate the development process of infection, inflammation or other complications. In this embodiment, the lung CT image is processed by denoising, smoothing, enhancing and filtering background effects to improve image quality and recognizability. If there is noise in the lung CT image, it will affect the observation and diagnosis of key lesions. The noise in the image is reduced by a denoising algorithm (wavelet transform) and important structural information is retained. Then, by smoothing (median filtering), unnecessary details in the image are reduced, the boundary of the lesion area is highlighted, and the lesion is clearer. Then, by contrast enhancement (Laplacian operator enhancement), the contrast between the lesion and normal tissue in the image is improved, so that the lesion area is easier to identify. Finally, a segmentation neural network is used to identify the lung CT image and background area in the complete lung CT image (including part of the background area) obtained by segmentation CT scanning (Computed Tomography). Therefore, the collected lung CT images are stored in a standardized manner to ensure the traceability of the image sequence, and the lung CT images and related clinical information (such as infection markers and patient symptoms) at each sampling moment are annotated for subsequent analysis and guidance.

[0046] It should be further explained that wavelet transform, median filtering, Laplace operator enhancement and segmentation neural network are all well-known technologies, and the specific methods are not introduced here. Among them, the relevant contents of the segmentation neural network are as follows: the segmentation neural network used in this embodiment is the Mask R-CNN neural network, and the data set used is a complete lung CT image data set obtained by CT scanning. Among them, Mask R-CNN is a well-known technology, and the specific method is not introduced here. The full name of Mask R-CNN in Chinese is "Mask Region Convolutional Neural Network", and the full name in English is "Mask Region-based Convolutional Neural Network". The pixels that need to be segmented are divided into 2 categories, that is, the labeling process corresponding to the training set is: a single-channel semantic label, the corresponding position pixel belongs to the lung CT image is marked as 0, and the label belongs to the background area is 1. The task of the network is classification, so the loss function used is the cross entropy loss function. Thus, the lung CT image and the background area in the complete lung CT image are obtained by segmenting the neural network. Therefore, through the above processing, the clarity and contrast of the lung CT image are improved, so that nursing staff can more clearly identify early signs of lung infection or other complications. Combined with the lesion information in the image, the nursing plan (such as position adjustment, breathing training, and anti-infection measures) can be adjusted in a timely manner to effectively reduce the incidence of complications such as postoperative infection.

[0047] Module 102: Pulmonary consolidation information analysis module.

[0048] This module is used to divide the lung CT images into several lung clusters, and determine the consolidation information value of the lung cluster according to the number of pixels and the size of the gray value in the lung cluster.

[0049] It should be noted that in this embodiment, by identifying and analyzing the lung CT images of lung cancer patients at multiple sampling moments after surgery, the possibility of postoperative infection complications of the patients reflected by the images is obtained, and relevant auxiliary nursing guidance suggestions are provided based on the possibility. Therefore, it is necessary to first perform detection and analysis of lung CT images, and extract and analyze indicators such as infection rate based on the image content.

[0050] Preferably, in one embodiment of the present invention, the method for obtaining the consolidation information value of each lung cluster includes:

[0051] For the lung CT image at any sampling time, the absolute value of the difference between the grayscale values ​​of any two pixels is taken as the clustering distance of the arbitrary two pixels. The iterative self-organizing clustering algorithm is used to perform clustering operations on all pixels to obtain several lung clusters.

[0052] It should be noted that the iterative self-organizing clustering algorithm is a well-known technology, and the specific method is not introduced here. Each lung cluster obtained has the same visual feature information, which means that each lung cluster has the same or similar lung health information. Because clustering is performed by gray value, parts with the same gray value or gray distribution information will be considered as a lung cluster. For example, after lung cancer surgery, pneumonia caused by reduced immunity of patients will appear in the image at this time, and high-density areas will appear as consolidation or spots, etc., then such consolidation areas will be considered as a lung cluster.

[0053] It should be further explained that each lung cluster has a consolidation information, which means that the lung cluster has the performance characteristics of possible lung inflammation and other complications. The more obvious the performance characteristics are, for example, the more severe the pneumonia infection caused by reduced immunity after lung cancer surgery, the more obvious the corresponding consolidation characteristics are, and the more sufficient the consolidation information is. Specifically, the consolidation variable of each lung cluster is analyzed, and consolidation refers to the obvious high-density area in the lung, which usually indicates inflammation, hematoma or tumor problems in the lung tissue. For lung infections (such as pneumonia), consolidation is usually a manifestation of infection or local inflammation. In lung CT images, consolidation is manifested as areas with higher density than normal lung tissue, and such areas are signs of pneumonia lesions.

[0054] For The first sampling time in the lung CT image lung clusters, and obtain the The mean gray value of all pixels in the lung cluster , calculate the The gray value of each pixel in the lung cluster is The absolute value of the difference between The grayscale values ​​of all pixels in the lung clusters are The sum of the absolute values ​​of the differences The inverse normalized value of The inverse normalized value of The product of The sum of the number of pixels in each lung cluster The normalized value of The consolidation information value of the lung clusters.

[0055] It should be noted that: in this embodiment, To present and value The inverse proportional relationship and normalization processing of the implementation can be set according to the actual situation. is an exponential function with a natural constant as the base, Linear normalization function for sum value Normalize the data values ​​to interval, taking this as an example. Since the higher the pixel grayscale mean, the higher the possibility of the corresponding lesion, because in the CT images of pneumonia, patchy or plaque images are common, especially in pneumonia with a large infection range or long-term untreated pneumonia, the CT images will show relatively uniform lung white spots. The distribution, size and morphology of these spots are usually manifested as white highlights and irregularities. Therefore, the more total number of pixels in the lung cluster, and the higher its grayscale mean, the higher the corresponding consolidation information, that is, the higher the possibility of pneumonia. Therefore, by analyzing the pixel distribution characteristics and pixel quantity in the lung cluster, the density information in the cluster is determined, and then the consolidation information value is obtained. By weighted summing the pixel quantity information and grayscale information in the cluster, the density information of the cluster is obtained and used as the consolidation information. When the total amount of pixels in the cluster is larger, the corresponding cluster density is higher, and accordingly, the grayscale value in the cluster is larger, and the grayscale distribution is uniform, that is, The larger the sum The smaller the value (in terms of sum The inverse normalized value of The higher the weight of the cluster, the higher the corresponding cluster density is, thereby obtaining the real change information value of each lung cluster. The higher the real change information value is, the higher the corresponding regional real variable is, and accordingly, the higher the infection rate of the patient is.

[0056] Module 103: Pulmonary consolidation result determination module.

[0057] This module is used to screen out several suspected infected lung clusters according to the size of the real change information value; record the connected area formed by adjacent pixel points in each suspected infected lung cluster as the lung cluster area; determine the regularity of the lung cluster area boundary according to the distribution of pixel points on the lung cluster area boundary; adjust the real change information value of the suspected infected lung cluster according to the regularity, and determine the real change result of the suspected infected lung cluster.

[0058] It should be noted that when the consolidation information value is higher, the patient's risk of pneumonia infection is higher. When the pneumonia infection is more serious, such as lung abscess or pneumothorax, lung CT images will show lung cavities or abscesses, usually manifested as air-filled cavities and irregular edges. Combined with the characteristics of spot lesions in the lung clusters, the irregularity of such edges can more completely reflect the patient's infection risk.

[0059] Preferably, in one embodiment of the present invention, the method for obtaining the consolidation result of each suspected infected lung cluster includes:

[0060] For a lung CT image at any sampling time, when the consolidation information value of any lung cluster is greater than a preset infection threshold, the any lung cluster is recorded as a suspected infected lung cluster.

[0061] It should be noted that: in this embodiment, the preset infection threshold is 0.9, which is used as an example for description. When the consolidation information value is greater than the preset infection threshold, it is considered that the lung cluster has a high possibility of lung infection, and it is necessary to further analyze the shape characteristics of the lung cluster to determine the corresponding cluster consolidation result.

[0062] In the The first sampling time in the lung CT image In the suspected infected lung cluster, the connected area formed by adjacent pixels is recorded as the lung cluster area.

[0063] It should be noted that: in this embodiment, the lung clusters are obtained by clustering based on the differences between the gray values ​​of the pixels, so each lung cluster may be composed of multiple connected areas. For example, a suspected infected lung cluster consists of three lung cluster areas, which means that there may be three lesion spots with relatively consistent gray values, or relatively regular tissue parts, in the suspected infected lung cluster. The preset area threshold is 10. If the area of ​​the lung cluster area is less than 10, the lung cluster area will not be used as a lung cluster area for subsequent analysis. This example is used for description. The lung cluster area with a small area is probably composed of interference points, not infection areas, so it is not analyzed.

[0064] For The first sampling moment in the lung CT image The first cluster of suspected lung infection lung cluster regions, and obtain the The center pixel of the lung cluster region is The Euclidean distance of each pixel on the boundary of the lung cluster region is calculated. The center pixel of the lung cluster region is The mean of the Euclidean distances of all pixels on the boundary of the lung cluster region , with the The central pixel of the lung cluster area is a circle. As the radius, construct the The circular area corresponding to the lung cluster area.

[0065] What needs to be explained is: the schematic diagram of the lung cluster area and the corresponding circular area, such as Figure 3 shown. Figure 3 In the figure, any pixel point on the boundary of the lung cluster area is taken as a sampling point, and each dotted line is the distance from the center of the lung cluster area to each sampling point. According to the above set rules, the rule judgment circle (circular area) of the lung cluster area can be obtained.

[0066] Get the The boundary of the lung cluster region is The number of overlapping pixels at the circular region boundaries corresponding to the lung cluster regions , calculate the The center pixel of the lung cluster region is The Euclidean distance of each pixel on the boundary of the lung cluster region is The absolute value of the difference between The center pixel of the lung cluster region is The Euclidean distance of all pixels on the boundary of the lung cluster region is The sum of the absolute values ​​of the differences The inverse normalized value of The inverse normalized value of The product of The normalized value of Regularity of the boundaries of the lung cluster regions.

[0067] It should be noted that: in this embodiment, To present and value The inverse proportional relationship and normalization processing of the implementation can be set according to the actual situation. is an exponential function with a natural constant as the base, Linear normalization function pair product Normalize the data values ​​to The above The larger the value is, the closer the lung cluster area boundary is to a circle and the more regular it is, and the higher the corresponding lung health level is, that is, the lower the possibility of infection is. The Euclidean distance between the central pixel of the lung cluster area and each pixel on the boundary of the lung cluster area is The smaller the difference, the more regular the boundary of the lung cluster area is, and the more likely it is to be close to the normal tissue part of the patient's lung, so the possibility of infection in the patient is lower.

[0068] The first The first sampling time in the lung CT image The mean of the regularity of the boundaries of all lung cluster regions within the suspected infected lung cluster is taken as the Regularity of clusters of suspected infected lungs.

[0069] The first The consolidation information value of the suspected infected lung cluster is the same as that of the Ratio of regularity of clusters of suspected infected lungs The normalized value of Consolidation results of a suspected infected lung cluster.

[0070] It should be noted that: in this embodiment, Linear normalization function comparison value Normalize the data values ​​to The ratio is within the range, and this is used as an example to describe. The larger the value, the more obvious the corresponding consolidation result and the higher the infection rate. When the consolidation results of all suspected infected lung clusters are larger, it means that the infection rate of lung cancer patients is higher.

[0071] Module 104: Nursing guidance recommendation determination module.

[0072] This module is used to determine the infection rate of lung cancer patients at each sampling moment according to the size of the solid change result; determine the diffusion properties of lung cancer patients according to the difference in infection rates of lung cancer patients at adjacent sampling moments; and provide different levels of nursing guidance suggestions to lung cancer patients according to the size of the diffusion properties.

[0073] Preferably, in one embodiment of the present invention, the method for acquiring the diffusion attribute of a lung cancer patient includes:

[0074] The first The mean of the consolidation results of all suspected infected lung clusters in the lung CT images at the sampling time is taken as the The infection rate of lung cancer patients at each sampling time.

[0075] What needs to be explained is: If there is no lung cluster area in the lung CT image at the sampling time, then The infection rate of lung cancer patients at the sampling time is 0, which is described as an example.

[0076] According to the above method, the infection rate of lung cancer patients at each sampling time is obtained.

[0077] It should be noted that in order to achieve efficient and accurate nursing care and prevent complications, it is necessary to analyze the infection rate of patients at different sampling times and determine the corresponding diffusion properties based on the infection rate. When the infection rate shows an increasing trend over time, the patient's diffusion property is higher, that is, after lung cancer surgery, the patient's own immunity decreases, leading to complications such as pneumonia. At this time, timely guidance and suggestions on relevant nursing care are needed.

[0078] Get the The infection rate of lung cancer patients at the sampling time minus the infection rate of The difference in infection rates of lung cancer patients at different sampling times ,like , then assign the Sampling time to The label of the sampling moment is 0. , then assign the Sampling time to The label of each sampling moment is 1. , then assign the Sampling time to The label of the sampling moment is -1, and the sum of the labels of all adjacent sampling moments is The normalized value of is recorded as the diffusion attribute of lung cancer patients.

[0079] It should be noted that: in this embodiment, Linear normalization function for sum value Normalize the data values ​​to When the infection rate of patients shows an increasing trend, let the label be 1. The larger the sum of all labels, the higher the diffusion attribute value, and the higher the necessity of nursing to prevent complications.

[0080] When the diffusion attribute of a lung cancer patient is greater than a preset diffusion threshold, the lung cancer patient is given first-level nursing guidance suggestions; when the diffusion attribute of a lung cancer patient is less than or equal to the preset diffusion threshold, the lung cancer patient is given second-level nursing guidance suggestions.

[0081] It should be noted that in this embodiment, the preset diffusion threshold is 0.8, which is used as an example for description. When the diffusion attribute is high, it is considered that the patient has a high possibility of spreading infection, and the first-level nursing guidance is given. When the diffusion attribute is low, it is considered that the patient has a low possibility of spreading infection, and the second-level nursing guidance is given. Among them, the nursing guidance acquisition flow chart is as follows: Figure 2 shown.

[0082] It should be further explained that the first-level nursing guidance recommendations include: breathing pattern adjustment, body position management, monitoring and drug use. Among them, breathing pattern adjustment includes: deep breathing training, encouraging patients to take deep breaths, helping alveolar expansion, clearing lung secretions, and reducing the risk of lung collapse. Effective cough training, guiding patients to use appropriate coughing techniques to eliminate sputum, avoid sputum accumulation in the lungs, and reduce the risk of infection. Body position management includes: intermittently encouraging patients to sit or semi-recumbent, which helps lung expansion, relieves pressure on the lower lobe, and promotes gas exchange. Perform regular body position changes, change body positions every hour, avoid long-term pressure on certain areas of the lungs, and reduce the occurrence of static areas of the lungs. Maintain a semi-recumbent or sitting position. For patients with a higher risk of infection, avoid long-term bed rest. Maintaining a sitting or semi-recumbent position at a certain angle can help reduce airway occlusion and pulmonary edema. Monitoring and drug use include: regular monitoring of physical signs (including body temperature, respiratory rate, and blood oxygen saturation, etc.), and timely detection of abnormalities. Drug management, early use of antibiotics or antiviral drugs for preventive treatment, to reduce the risk of secondary infection. Review images and perform CT scans regularly to monitor changes in the lungs and determine whether pneumonia is getting worse.

[0083] Secondary nursing guidance recommendations include: breathing pattern adjustment, body position management, and routine monitoring and care. Among them, breathing pattern adjustment includes: moderate breathing training, and it is recommended to conduct mild breathing training and cough training to maintain lung vitality and prevent mucus accumulation. Maintain normal breathing frequency and depth, and encourage patients not to breathe too deeply, but to maintain normal breathing depth and frequency to avoid shallow breathing. Body position management includes: normal body position change, encourage patients to change their body position regularly, avoid maintaining the same posture for a long time, but do not need to adjust the body position too frequently. Avoid long-term bed rest, and patients can carry out moderate activities to keep their bodies flexible and promote lung function. Routine monitoring and care include: mild monitoring, regular monitoring of body temperature and breathing rate, etc., to ensure that there are no abnormalities. Strengthen respiratory care, encourage patients to drink water, keep the respiratory tract moist, and avoid respiratory infections caused by dryness.

[0084] So far, the present invention is completed.

[0085] In summary, in an embodiment of the present invention, a lung CT image of a lung cancer patient at each sampling moment after surgery is obtained, and several lung clusters are divided, and the consolidation information value of the lung cluster is determined to screen out several suspected infected lung clusters, and the connected area composed of adjacent pixels in each suspected infected lung cluster is recorded as a lung cluster area, and the regularity of the lung cluster area boundary is determined. The consolidation information value is adjusted according to the regularity, and the consolidation result of the suspected infected lung cluster is determined to determine the infection rate of the lung cancer patient at each sampling moment, thereby obtaining the diffusion properties of the lung cancer patient, and giving different levels of nursing guidance suggestions to the lung cancer patient. The present invention obtains corresponding infection prediction diffusion results and corresponding nursing guidance suggestions by analyzing the lung CT images of the patient at different periods, thereby reducing the risk of postoperative complications and improving the accuracy of nursing.

[0086] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. An intelligent image-assisted guidance system for postoperative nursing of lung cancer, characterized in that: The system includes the following modules: Lung CT image acquisition module: used to obtain lung CT images of lung cancer patients at each sampling moment after surgery; Lung consolidation information analysis module: used to divide the lung CT images into several lung clusters, and determine the consolidation information value of the lung cluster according to the number of pixels and the size of the gray value in the lung cluster; Lung consolidation result determination module: used to screen out a number of suspected infected lung clusters according to the size of the consolidation information value; record the connected area formed by adjacent pixels in each suspected infected lung cluster as a lung cluster area; determine the regularity of the lung cluster area boundary according to the distribution of pixels on the lung cluster area boundary; adjust the consolidation information value of the suspected infected lung cluster according to the regularity, and determine the consolidation result of the suspected infected lung cluster; Nursing guidance recommendation determination module: used to determine the infection rate of lung cancer patients at each sampling time according to the size of the solid change result; determine the diffusion property of lung cancer patients according to the difference in infection rate of lung cancer patients at adjacent sampling times; and give different levels of nursing guidance recommendations to lung cancer patients according to the size of the diffusion property.

2. According to claim 1, an intelligent image-assisted guidance system for postoperative nursing of lung cancer is characterized in that: Determining the consolidation information value of the lung cluster includes: In any one lung cluster, the mean of the grayscale values ​​of all pixels is obtained, the inversely proportional normalized value of the sum of the absolute values ​​of the differences between the grayscale values ​​of all pixels and the mean is obtained, the product of the inversely proportional normalized value and the mean is calculated, and the product and the normalized value of the sum of the number of pixels in the any one lung cluster are used as the real change information value of the any one lung cluster.

3. According to claim 1, an intelligent image-assisted guidance system for postoperative nursing of lung cancer is characterized in that: The screening of several suspected lung infection clusters includes: When the consolidation information value of any lung cluster is greater than a preset infection threshold, the any lung cluster is recorded as a suspected infected lung cluster.

4. According to claim 1, an intelligent image-assisted guidance system for postoperative nursing of lung cancer is characterized in that: The regularity of determining the lung cluster region boundaries includes: For any lung cluster region, obtain the mean of the Euclidean distances between the central pixel of the lung cluster region and all the pixel points on the boundary of the lung cluster region, and construct a circular region with the central pixel of the lung cluster region as the dot and the mean as the radius; The regularity of the lung cluster region boundary is determined based on the overlap between the lung cluster region boundary and the circular region boundary and the Euclidean distance from the central pixel point of the lung cluster region to the lung cluster region boundary.

5. According to claim 4, an intelligent image-assisted guidance system for postoperative nursing of lung cancer is characterized in that: The determining of the regularity of the lung cluster region boundary according to the overlap between the lung cluster region boundary and the circular region boundary and the Euclidean distance from the central pixel point of the lung cluster region to the lung cluster region boundary includes: The number of overlapping pixels between the boundary of the lung cluster region and the boundary of the circular region is obtained, and an inversely proportional normalized value of the sum of the absolute values ​​of the differences between the Euclidean distances between the central pixel of the lung cluster region and all the pixels on the boundary of the lung cluster region and the mean is calculated; and a normalized value of the product of the inversely proportional normalized value and the number of overlapping pixels is used as the regularity of the boundary of the lung cluster region.

6. According to claim 1, an intelligent image-assisted guidance system for postoperative nursing of lung cancer is characterized in that: The results of the consolidation of suspected infected lung clusters include: Taking the average of the regularities of all lung cluster region boundaries within any suspected infected lung cluster as the regularity of the suspected infected lung cluster; The normalized value of the ratio of the consolidation information value to the regularity of the suspected infected lung cluster is used as the consolidation result of the suspected infected lung cluster.

7. The intelligent image-assisted guidance system for postoperative nursing care of lung cancer according to claim 1, characterized in that: Determining the infection rate of lung cancer patients at each sampling time includes: The mean of the consolidation results of all suspected infected lung clusters in the lung CT images at any sampling time is taken as the infection rate of lung cancer patients at the sampling time.

8. The intelligent image-assisted guidance system for postoperative nursing care of lung cancer according to claim 1, characterized in that: Determining the diffusion attribute of the lung cancer patient includes: Get the The sampling time and The difference in infection rates of lung cancer patients at different sampling times ,like , then assign the Sampling time to The label of the sampling moment is 0. , then assign the Sampling time to The label of each sampling moment is 1. , then assign the Sampling time to The label of the sampling moment is -1; Determine the diffusion properties of lung cancer patients based on the labels of all adjacent sampling moments.

9. The intelligent image-assisted guidance system for postoperative nursing care of lung cancer according to claim 8, characterized in that: Determining the diffusion attribute of the lung cancer patient according to the labels of all adjacent sampling moments includes: The normalized value of the sum of the labels at all adjacent sampling moments is recorded as the diffusion attribute of the lung cancer patient.

10. The intelligent image-assisted guidance system for postoperative nursing of lung cancer according to claim 1, characterized in that: According to the size of the diffusion attribute, the different levels of nursing guidance suggestions for lung cancer patients include: When the diffusion attribute of a lung cancer patient is greater than a preset diffusion threshold, the lung cancer patient is given first-level nursing guidance suggestions; when the diffusion attribute of a lung cancer patient is less than or equal to the preset diffusion threshold, the lung cancer patient is given second-level nursing guidance suggestions.

Images