Difficulty airway management method based on CT image analysis of oral cavity head and neck tumor patient
Through CT imaging analysis and 3D reconstruction technology, combined with LIME algorithm and multimodal data, we can identify and evaluate the difficult airways of patients with oral head and neck tumors, generate a three-dimensional airway model and perform virtual assisted intubation, solving the difficulty of accurate management of the existing technology. Complex airway problems achieve more efficient and safe airway management.
Patent Information
- Application Number
- CN202411501547.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-25
- Publication Date
- 2025-02-18
AI Technical Summary
The airway structure of oral head and neck tumors changes significantly due to tumor compression or erosion. The prior art is difficult to accurately reflect these complex changes, resulting in increased difficulty and risk of difficult airway management.
Through CT image analysis technology and 3D reconstruction technology, a multimodal difficult airway intelligent evaluation model based on LIME algorithm is constructed. Combining the patient's CT image and multiple clinical features, key feature areas are identified, a three-dimensional airway model is generated, and a three-dimensional airway model is generated through 3D Print and virtual endoscopic techniques perform pre-cannula exercises and real-time assisted cannula.
It improves the identification accuracy and operation success rate of difficult airways, provides personalized management strategies, and significantly improves the safety and management success rate of perioperative patients.
Smart Images
Figure CN119454061A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of difficult airway management methods, and in particular to a difficult airway management method based on CT image analysis of patients with oral head and neck tumors. Background Art
[0002] Patients with oral head and neck tumors often have anatomical abnormalities and airway compression or erosion, which makes traditional airway management methods face many challenges in these patients. Under the compression or erosion of the tumor, the patient's airway structure often changes significantly, and traditional imaging assessment methods are difficult to accurately reflect these complex changes. Existing difficult airway management usually relies on the doctor's experience and preoperative static images (such as CT images) for evaluation, but these methods are difficult to accurately describe the complex airway anatomical changes in tumor patients. Static images can only provide anatomical information at a specific time point, making the prediction and management of difficult airways more complicated. The airway may change dynamically during surgery due to factors such as tumor resection and body position changes. Existing static imaging assessment methods cannot provide real-time airway monitoring and adjustment, which increases the difficulty and risk of intraoperative airway management. Existing technologies usually adopt a unified standardized management plan and lack personalized response measures for individual patient differences. The lesion site, tumor size and shape, and impact on the airway of each oral head and neck tumor patient are different. Standardized plans cannot effectively deal with complex individual differences, which can easily lead to intraoperative intubation failure or serious complications.
[0003] Therefore, it is necessary to propose a new difficult airway management method based on CT image analysis of patients with oral head and neck tumors. Summary of the invention
[0004] The present invention provides a difficult airway management method based on CT image analysis of patients with oral head and neck tumors. Through CT image analysis technology and 3D reconstruction technology, a more accurate and efficient difficult airway management solution is provided, which improves the accuracy of difficult airway identification and the success rate of operation.
[0005] The present invention provides a difficult airway management method based on CT image analysis of patients with oral head and neck tumors, comprising:
[0006] Constructing a multimodal difficult airway intelligent assessment model based on the LIME algorithm, wherein the multimodal difficult airway intelligent assessment model based on the LIME algorithm is used to combine the patient's CT image with multiple clinical characteristics to identify key feature areas in the CT image that are highly correlated with the occurrence of difficult airway;
[0007] Based on the key feature areas, a three-dimensional airway model is generated using airway reconstruction technology, and a solid physical model is generated using 3D printing technology, wherein the solid physical model is used for pre-intubation practice;
[0008] The virtual endoscope technology is used to display the key feature areas in real time during the intubation process, providing visual reference to assist intubation.
[0009] Optionally, the constructing of a multimodal difficult airway intelligent assessment model based on the LIME algorithm comprises:
[0010] Collect CT images and various clinical features of patients with head and neck tumors and standardize the image data;
[0011] The Transformer architecture is used to combine the patient's clinical characteristics and CT images to perform multi-scale feature analysis and extract linear and nonlinear features;
[0012] The LIME algorithm technology is used to interpret the model, and key feature areas that are highly correlated with the occurrence of difficult airway are identified from the extracted linear features and the nonlinear features;
[0013] A secondary analysis is performed on the identified key feature areas to extract features that are highly correlated with airway pathological changes, and the features are condensed to form feature rules.
[0014] Optionally, the difficult airway management method based on CT image analysis of patients with oral head and neck tumors also includes:
[0015] Several high-risk patients who were determined to have difficult airway were randomly divided into a conventional difficult airway management group and a digital personalized management group. The conventional airway management group did not adopt the difficult airway management method based on CT image analysis of patients with oral head and neck tumors, and the digital personalized management group adopted the difficult airway management method based on CT image analysis of patients with oral head and neck tumors.
[0016] The intubation success rate, patient blood oxygen saturation changes, and operation time of the conventional airway management group and the digital personalized management group were compared to evaluate the effectiveness of the difficult airway management method based on CT image analysis of patients with oral head and neck tumors.
[0017] Optionally, the key feature areas include the position of the hyoid bone and the shape of the tongue.
[0018] Optionally, the features that are highly correlated with airway pathological changes include texture features and morphological features.
[0019] Optionally, the multiple clinical characteristics include respiratory parameters, medical history, and patient vital signs.
[0020] Optionally, the difficult airway management method based on CT image analysis of patients with oral head and neck tumors also includes:
[0021] Use blockchain technology to record and track patient data and analysis results, ensure the authenticity and security of the data, and prevent data tampering during transmission and sharing.
[0022] Optionally, the difficult airway management method based on CT image analysis of patients with oral head and neck tumors also includes:
[0023] The multimodal difficult airway intelligent assessment model based on the LIME algorithm is deployed on a dedicated server or mobile device in the operating room, and combined with cloud computing for data analysis and model optimization to reduce latency and improve computing efficiency.
[0024] Optionally, the constructing of a multimodal difficult airway intelligent assessment model based on the LIME algorithm further includes:
[0025] Magnetic resonance imaging and positron emission tomography image data are integrated to enhance the assessment capability of the multimodal difficult airway intelligent assessment model based on the LIME algorithm.
[0026] Optionally, the difficult airway management method based on CT image analysis of patients with oral head and neck tumors also includes:
[0027] Using AR or VR technology, real-time three-dimensional views and dynamic feedback are provided during the intubation operation.
[0028] Compared with the prior art, the technical solution of the present invention has the following beneficial effects:
[0029] 1. The technical solution of the present invention integrates CT images with multiple clinical features (such as patient signs, medical history, respiratory parameters, etc.) to create a multimodal difficult airway intelligent assessment model based on the LIME algorithm. It integrates and analyzes multiple data sources through deep learning algorithms (such as the Transformer architecture), thereby improving the accuracy and comprehensiveness of difficult airway prediction.
[0030] 2. The present invention introduces the LIME algorithm to enhance the interpretability of the model. Based on the traditional deep learning model, the present invention introduces the LIME (Local Interpretable Model-Agnostic Explanations) algorithm to improve the interpretability and transparency of the model. The LIME algorithm enables clinicians to understand and trust the prediction results of the model by identifying key feature areas in CT images;
[0031] 3. This invention introduces 3D printing technology and virtual endoscopy technology into the airway management process, and by converting the key areas of CT image analysis into a physical airway model and intraoperative virtual visual reference, it enhances the clinician's perception of difficult airways and operational guidance;
[0032] 4. The present invention provides a management strategy that can be tailored according to the patient's specific anatomical structure, tumor location and lesion characteristics. This method significantly improves the safety and success rate of difficult airway management in perioperative patients. BRIEF DESCRIPTION OF THE DRAWINGS
[0033] In order to more clearly illustrate the embodiments of the present application 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 application. For ordinary technicians in this field, other drawings can be obtained based on the structures shown in these drawings without paying any creative work.
[0034] Figure 1 This is a flow chart of a difficult airway management method based on CT image analysis of oral head and neck tumor patients according to an embodiment of the present invention;
[0035] Figure 2 A schematic diagram of the construction process of a multimodal difficult airway intelligent assessment model based on the LIME algorithm according to an embodiment of the present invention;
[0036] Figure 3 A schematic diagram of a 3D printing technology for pre-intubation training and virtual endoscopy technology for assisting intubation according to an embodiment of the present invention;
[0037] Figure 4 The figure is a schematic diagram of the effectiveness verification process of a difficult airway management method based on CT image analysis of oral head and neck tumor patients according to an embodiment of the present invention. DETAILED DESCRIPTION
[0038] The following is a further detailed description of the difficult airway management method based on CT image analysis of patients with oral head and neck tumors proposed by the present invention in conjunction with the accompanying drawings and specific embodiments. According to the following description, the advantages and features of the present invention will be clearer. It should be noted that the drawings are in a very simplified form and use non-precise proportions, which are only used to conveniently and clearly assist in explaining the purpose of the embodiments of the present invention. In order to make the purposes, features and advantages of the present invention more obvious and easy to understand, please refer to the accompanying drawings. It should be noted that the structures, proportions, sizes, etc. illustrated in the drawings of this specification are only used to match the contents disclosed in the specification for people familiar with this technology to understand and read, and are not used to limit the limiting conditions for the implementation of the present invention, so they have no technical substantive significance. Any structural modification, change in proportional relationship or adjustment of size, without affecting the effects that can be produced by the present invention and the purposes that can be achieved, should still fall within the scope of the technical content disclosed by the present invention.
[0039] See also Figure 1 , Figure 1The present invention is a flowchart of a difficult airway management method based on CT image analysis of oral head and neck tumor patients according to an embodiment of the present invention, comprising:
[0040] Step S1: constructing a multimodal difficult airway intelligent assessment model based on the LIME (Local Interpretable Model-Agnostic Explanations) algorithm, wherein the multimodal difficult airway intelligent assessment model based on the LIME algorithm is used to combine the patient's CT image and multiple clinical characteristics to identify key feature areas in the CT image that are highly correlated with the occurrence of difficult airway;
[0041] Step S2: Based on the key feature area, a three-dimensional airway model is generated using airway reconstruction technology, and a solid physical model is generated using 3D printing technology, wherein the solid physical model is used for pre-intubation practice;
[0042] Step S3: Using virtual endoscopy technology to display the key feature area in real time during the intubation process, providing visual reference to assist intubation.
[0043] This embodiment integrates CT images and multiple clinical features to perform intelligent assessment of difficult airways. By combining the patient's clinical features and CT images, the model can more accurately identify and predict difficult airways. The LIME algorithm helps identify key feature areas in CT images, which are highly correlated with the occurrence of difficult airways, allowing clinicians to understand the reasons behind the predictions and increase their trust in the model. The key feature areas predicted by the model are converted into visual physical and virtual models to assist clinicians in preoperative planning and intraoperative operations. 3D printing technology generates physical airway models for intubation practice, and virtual endoscopy technology provides real-time visual reference during surgery. The combination of these technologies helps to improve the success rate and safety of difficult airway management.
[0044] See also Figure 2 , Figure 2 The following is a schematic diagram of a construction process of a multimodal difficult airway intelligent assessment model based on the LIME algorithm according to an embodiment of the present invention, including:
[0045] Collect CT images and various clinical features of patients with head and neck tumors, and standardize the image data; a database including a large number of CT image data, clinical indicators and related records of patients with head and neck tumors is the basis for model training and optimization. Provide sufficient data support for the model to ensure wide coverage of different patient types, so as to improve the generalization ability and clinical applicability of the model;
[0046] The Transformer architecture is used to combine the patient's clinical characteristics and CT images to perform multi-scale feature analysis and extract linear and nonlinear features;
[0047] The LIME algorithm technology is used to interpret the model, and key feature areas that are highly correlated with the occurrence of difficult airways are identified from the extracted linear features and nonlinear features; the LIME algorithm is used to improve the interpretability and transparency of the model, and help identify key feature areas in CT images that are highly correlated with the occurrence of difficult airways. By explaining and annotating the prediction results of the model, clinicians can understand the reasons behind the prediction, improve their trust in the model, and facilitate its application in clinical practice;
[0048] A secondary analysis is performed on the identified key feature areas to extract features that are highly correlated with airway pathological changes, and feature condensation is performed to form feature rules. In some embodiments, the key feature areas include hyoid bone position and tongue morphology. In some embodiments, the features that are highly correlated with airway pathological changes include texture features and morphological features.
[0049] In other embodiments, based on the LIME algorithm, explanatory algorithms such as SHAP (Shapley Additive Explanations) algorithm and Grad-CAM (Gradient-weighted Class Activation Mapping) are further introduced to enhance the multi-level explanatory ability of the model. When faced with more complex clinical situations or multimodal data sources, the combined use of different explanatory algorithms can more comprehensively reveal the reasons behind the model's decisions and improve the credibility of the model.
[0050] In some other embodiments, in addition to the Transformer architecture, other deep learning models (such as convolutional neural network CNN, recurrent neural network RNN, graph convolutional network GCN) can be tried to predict and evaluate difficult airways. Different models may provide better performance in specific anatomical structure analysis (such as local feature extraction in images) or time series data (such as respiratory cycle monitoring).
[0051] See also Figure 3 , Figure 3This is a schematic diagram of the use of 3D printing technology for pre-intubation practice and virtual endoscopy technology to assist intubation in one embodiment of the present invention. Through 3D printing technology, clinicians can conduct physical operation training on individual patients' airway structures before surgery, so as to become familiar with the anatomical characteristics of difficult airways. Virtual endoscopy technology provides real-time visual reference during surgery to improve the success rate of intubation and operational safety. This comprehensive support from preoperative to intraoperative significantly improves the overall efficiency and effectiveness of difficult airway management. The application of 3D printed models and virtual endoscopy technology not only helps to improve the operational skills of existing doctors, but can also be used as a teaching tool to assist medical education and clinical training, thereby cultivating more doctors with practical experience and skills in the field of airway management.
[0052] In some embodiments, the difficult airway management method based on CT image analysis of oral head and neck tumor patients further includes:
[0053] Several high-risk patients who were determined to have difficult airway were randomly divided into a conventional difficult airway management group and a digital personalized management group. The conventional airway management group did not adopt the difficult airway management method based on CT image analysis of patients with oral head and neck tumors, and the digital personalized management group adopted the difficult airway management method based on CT image analysis of patients with oral head and neck tumors.
[0054] The intubation success rate, patient blood oxygen saturation changes, and operation time of the conventional airway management group and the digital personalized management group were compared to evaluate the effectiveness of the difficult airway management method based on CT image analysis of patients with oral head and neck tumors.
[0055] See also Figure 4 , Figure 4 This is a schematic diagram of the effectiveness verification process of the difficult airway management method based on CT image analysis of patients with oral head and neck tumors according to an embodiment of the present invention; the clinical effect of the method is verified through randomized controlled trials, and indicators such as the success rate of intubation and changes in patient blood oxygen saturation are compared to ensure the wide applicability and operability of the method in different patient groups. The personalized difficult airway management strategy provided by the present invention is customized according to the patient's specific anatomical and pathological characteristics, avoiding the "one-size-fits-all" universal approach, significantly improving the safety and management success rate of perioperative patients; reducing the occurrence of perioperative accidents and the risk of complications, thereby reducing the need for secondary surgery or further medical intervention due to airway management failure, saving medical resources and costs.
[0056] In some embodiments, the multiple clinical features include respiratory parameters, medical history, and patient signs. This embodiment fuses CT image data with multiple clinical features (such as patient signs, medical history, respiratory parameters, etc.) to create a multimodal difficult airway intelligent assessment model. The accuracy and comprehensiveness of difficult airway prediction are improved by integrating and analyzing multiple data sources through deep learning algorithms (such as Transformer architecture).
[0057] The difficult airway management method based on CT image analysis of oral head and neck tumor patients also includes:
[0058] Use blockchain technology to record and track patient data and analysis results, ensure the authenticity and security of the data, and prevent data tampering during transmission and sharing.
[0059] In some embodiments, the difficult airway management method based on CT image analysis of oral head and neck tumor patients further includes:
[0060] The multimodal difficult airway intelligent assessment model based on the LIME algorithm is deployed on a dedicated server or mobile device in the operating room, and combined with cloud computing for data analysis and model optimization to reduce latency and improve computing efficiency. It is suitable for medical application scenarios in operating room environments or remote areas that require real-time data analysis and decision support.
[0061] In some embodiments, the constructing of a multimodal difficult airway intelligent assessment model based on the LIME algorithm further includes:
[0062] Magnetic resonance imaging and positron emission tomography image data are integrated to enhance the assessment capability of the multimodal difficult airway intelligent assessment model based on the LIME algorithm.
[0063] In the above embodiment, in addition to CT images, data sources of other imaging technologies, such as MRI (magnetic resonance imaging), ultrasound imaging or PET (positron emission tomography), etc., can also be integrated to further enhance the comprehensive assessment capability of airway structure. When a comprehensive assessment of complex anatomical structures or pathological conditions is required (such as tumor invasion of surrounding tissues), data from different imaging modes can provide more dimensional information and improve the accuracy and coverage of the assessment.
[0064] In some embodiments, the difficult airway management method based on CT image analysis of oral head and neck tumor patients further includes:
[0065] Using AR or VR technology, real-time 3D views and dynamic feedback are provided during the intubation process, helping doctors to perform intubation more accurately. During intraoperative or postoperative review, doctors can use VR / AR technology to more intuitively understand the details of the operation process and improve training effects.
[0066] In summary, the difficult airway management method based on CT image analysis of patients with oral head and neck tumors provided in this embodiment integrates CT image data with multiple clinical features to construct a multimodal difficult airway intelligent assessment model. The accuracy and comprehensiveness of difficult airway prediction are improved by integrating and analyzing multiple data sources through deep learning algorithms (such as Transformer architecture). The introduction of the LIME (Local Interpretable Model-Agnostic Explanations) algorithm improves the interpretability and transparency of the model. The LIME algorithm enables clinicians to understand and trust the prediction results of the model by identifying key feature areas in CT images. 3D printing technology and virtual endoscopy technology are introduced into the airway management process. By converting the key areas of CT image analysis into physical airway models and intraoperative virtual visual references, clinicians' perception and operational guidance of difficult airways are enhanced. This combined application of multiple technical means realizes all-round support from preoperative planning to intraoperative assistance. The management strategy can be tailored according to the patient's specific anatomical structure, tumor location and lesion characteristics. This method significantly improves the safety of perioperative patients and the success rate of management.
[0067] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.
[0068] In the description of the present invention, it should be understood that the terms "center", "height", "thickness", "up", "down", "vertical", "horizontal", "top", "bottom", "inside", "outside", "axial", "radial", "circumferential" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In the description of the present invention, unless otherwise specified, "plurality" means two or more.
[0069] In the description of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", and "fixed" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.
[0070] In the present invention, unless otherwise clearly specified and limited, a first feature being "above" or "below" a second feature may include that the first and second features are in direct contact, or may include that the first and second features are not in direct contact but are in contact through another feature between them. Moreover, a first feature being "above", "above" and "above" a second feature includes that the first feature is directly above and obliquely above the second feature, or simply indicates that the first feature is higher in level than the second feature. A first feature being "below", "below" and "below" a second feature includes that the first feature is directly below and obliquely below the second feature, or simply indicates that the first feature is lower in level than the second feature.
[0071] Although the content of the present invention has been described in detail through the above preferred embodiments, it should be appreciated that the above description should not be considered as a limitation of the present invention. After reading the above content, it will be apparent to those skilled in the art that various modifications and substitutions of the present invention will occur. Therefore, the protection scope of the present invention should be limited by the appended claims.
Claims
1. A difficult airway management method based on CT image analysis of patients with oral head and neck tumors, characterized in that: include: Constructing a multimodal difficult airway intelligent assessment model based on the LIME algorithm, wherein the multimodal difficult airway intelligent assessment model based on the LIME algorithm is used to combine the patient's CT image with multiple clinical characteristics to identify key feature areas in the CT image that are highly correlated with the occurrence of difficult airway; Based on the key feature areas, a three-dimensional airway model is generated using airway reconstruction technology, and a solid physical model is generated using 3D printing technology, wherein the solid physical model is used for pre-intubation practice; The virtual endoscope technology is used to display the key feature areas in real time during the intubation process, providing visual reference to assist intubation.
2. The difficult airway management method based on CT image analysis of oral head and neck tumor patients according to claim 1, characterized in that: The multimodal difficult airway intelligent assessment model based on the LIME algorithm is constructed, including: Collect CT images and various clinical features of patients with head and neck tumors and standardize the image data; The Transformer architecture is used to combine the patient's clinical characteristics and CT images to perform multi-scale feature analysis and extract linear and nonlinear features; The LIME algorithm technology is used to interpret the model, and key feature areas that are highly correlated with the occurrence of difficult airway are identified from the extracted linear features and the nonlinear features; A secondary analysis is performed on the identified key feature areas to extract features that are highly correlated with airway pathological changes, and the features are condensed to form feature rules.
3. The difficult airway management method based on CT image analysis of oral head and neck tumor patients according to claim 1, characterized in that: Also includes: Several high-risk patients who were determined to have difficult airway were randomly divided into a conventional difficult airway management group and a digital personalized management group. The conventional airway management group did not adopt the difficult airway management method based on CT image analysis of patients with oral head and neck tumors, and the digital personalized management group adopted the difficult airway management method based on CT image analysis of patients with oral head and neck tumors. The intubation success rate, patient blood oxygen saturation changes, and operation time of the conventional airway management group and the digital personalized management group were compared to evaluate the effectiveness of the difficult airway management method based on CT image analysis of patients with oral head and neck tumors.
4. The difficult airway management method based on CT image analysis of oral head and neck tumor patients according to claim 2, characterized in that: The key feature areas include the position of the hyoid bone and the shape of the tongue.
5. The difficult airway management method based on CT image analysis of oral head and neck tumor patients according to claim 2, characterized in that: The features that are highly correlated with airway pathological changes include texture features and morphological features.
6. The difficult airway management method based on CT image analysis of oral head and neck tumor patients according to claim 1, characterized in that: The multiple clinical characteristics include respiratory parameters, medical history, and patient vital signs.
7. The difficult airway management method based on CT image analysis of oral head and neck tumor patients according to claim 2, characterized in that: Also includes: Use blockchain technology to record and track patient data and analysis results, ensure the authenticity and security of the data, and prevent data tampering during transmission and sharing.
8. The difficult airway management method based on CT image analysis of oral head and neck tumor patients according to claim 2, characterized in that: Also includes: The multimodal difficult airway intelligent assessment model based on the LIME algorithm is deployed on a dedicated server or mobile device in the operating room, and combined with cloud computing for data analysis and model optimization to reduce latency and improve computing efficiency.
9. The difficult airway management method based on CT image analysis of oral head and neck tumor patients according to claim 2, characterized in that: The method of constructing a multimodal difficult airway intelligent assessment model based on the LIME algorithm further includes: Magnetic resonance imaging and positron emission tomography image data are integrated to enhance the assessment capability of the multimodal difficult airway intelligent assessment model based on the LIME algorithm.
10. The difficult airway management method based on CT image analysis of oral head and neck tumor patients according to claim 1, characterized in that: Also includes: Using AR or VR technology, real-time three-dimensional views and dynamic feedback are provided during the intubation operation.