Intelligentization-based oral tumor detection method and system

Abstract

The invention relates to the field of data processing, and provides an intelligent oral tumor detection method and system, and the method comprises the steps: collecting oral basic information of a to-be-detected patient, carrying out the hierarchical scanning to obtain hierarchical scanning data, obtaining an oral calibration signal through a series of processing, generating an oral feature set based on the oral calibration signal, and carrying out the detection of the oral tumor. Screening core parameters by combining a database, analyzing an oral cavity change mode, mining an oral cavity oncogenesis degree and a quantitative relation thereof, constructing an oral cavity oncogenesis evaluation system, determining a tumor suspected region and region characteristics, analyzing a development trend and a potential diffusion range, determining a pathological state and a state characteristic cluster, judging an oral cavity tumor risk level, and extracting key pathogenic factors; and finally generating an oral tumor detection report. The accuracy of oral tumor detection can be improved.

Description

Technical Field

[0001] The present invention relates to the field of data processing, and in particular to an oral tumor detection method and system based on intelligence. Background Art

[0002] In today's medical field, oral tumors are a type of disease that seriously threatens human health. Their early detection and accurate diagnosis are of vital importance to improving treatment effects and patient prognosis. The occurrence and development of oral tumors is a complex process, which brings huge challenges to clinical diagnosis and treatment.

[0003] At present, traditional oral tumor detection methods have many limitations. For example, routine oral examinations often rely on the doctor's subjective judgment and experience, which is prone to missed diagnosis and misdiagnosis. Although histopathological examination is the gold standard for diagnosis, it is traumatic, and the detection process is relatively cumbersome and time-consuming. In addition, the existing detection methods are not capable of identifying early micro-lesions, and it is difficult to effectively integrate and analyze these multi-source heterogeneous data, resulting in inaccurate diagnostic results. Therefore, there is an urgent need for an intelligent oral tumor detection method and system to improve the accuracy of oral tumor detection. Summary of the invention

[0004] The present invention provides an oral tumor detection method and system based on intelligence, the main purpose of which is to improve the accuracy of oral tumor detection.

[0005] Collecting basic oral information of the patient to be tested, performing layered scanning on the oral cavity of the patient to be tested based on the basic oral information to obtain layered scanning data, detecting an oral scanning signal corresponding to the layered scanning data, performing signal denoising on the oral scanning signal to obtain a denoised oral signal, and performing calibration processing on the denoised oral signal to obtain an oral calibration signal;

[0006] Based on the oral calibration signal, an oral feature set corresponding to the patient's oral cavity is generated; based on the oral feature set combined with a preset large-scale oral tumor database, core parameters indicating tumors are screened; and based on the core parameters, an oral change pattern corresponding to the oral cavity of the patient to be tested is analyzed;

[0007] Mining the degree of oral neoplasia under the oral change pattern, analyzing the quantitative relationship corresponding to the degree of oral neoplasia, combining the quantitative relationship with the oral state of the patient to be tested, and constructing an oral neoplasia evaluation system corresponding to the patient to be tested;

[0008] Based on the oral neoplasia assessment system, determining the suspected tumor area corresponding to the oral cavity of the patient to be tested, identifying the regional features corresponding to the suspected tumor area, analyzing the development trend of the suspected tumor area based on the regional features, and calculating the potential diffusion range corresponding to the suspected tumor area based on the development trend;

[0009] Based on the potential spread range, the pathological state corresponding to the oral cavity of the patient to be tested is determined, the state feature cluster corresponding to the pathological state is analyzed, and based on the state feature cluster, the oral tumor risk level corresponding to the patient to be tested is judged, and the key pathogenic factors in the oral tumor risk level are extracted. Based on the key pathogenic factors, an oral tumor detection report for the patient to be tested is generated.

[0010] Optionally, performing a layered scan on the oral cavity of the patient to be detected based on the oral basic information to obtain layered scan data includes:

[0011] Query the basic parameters corresponding to the oral basic information;

[0012] Based on the basic parameters, a preliminary scan is performed on the oral cavity of the patient to be tested to obtain preliminary scan oral cavity data;

[0013] Performing denoising on the initial oral scan data to obtain denoised initial oral scan data;

[0014] Performing hierarchical planning on the denoised initial scan oral data to obtain a hierarchical planning range;

[0015] Based on the hierarchical planning range, a hierarchical scan is performed on the oral cavity of the patient to be detected to obtain hierarchical scan data.

[0016] Optionally, generating an oral feature set corresponding to the patient's oral cavity based on the oral calibration signal includes:

[0017] Performing in-depth analysis on the oral calibration signal and extracting key characteristic parameters from the oral calibration signal;

[0018] Based on the key characteristic parameters, defining a characteristic identifier corresponding to the oral calibration signal;

[0019] Performing association analysis on the feature identifiers to obtain a feature association matrix;

[0020] Screening the core features in the feature association matrix;

[0021] Based on the core features, an oral feature set corresponding to the patient's oral cavity is generated.

[0022] Optionally, the step of combining the oral feature set with a preset large-scale oral tumor database to screen core parameters indicating tumors includes:

[0023] Performing feature mapping between the features in the oral feature set and the features in a preset large-scale oral tumor database to obtain mapped feature cases;

[0024] Calculating case similarity corresponding to the mapped feature case;

[0025] Based on the case similarity, determining the corresponding similar case set;

[0026] Performing frequency statistics on the features in the similar case set to obtain the frequency of feature occurrence;

[0027] Screening high-frequency features from the occurrence frequencies of the features; performing correlation analysis on the high-frequency features to obtain a feature correlation matrix;

[0028] Extracting strongly correlated feature combinations in the feature correlation matrix;

[0029] Performing a clinical evaluation on the strongly correlated feature combination to obtain an evaluation result;

[0030] Determine the core parameters indicating the presence of a tumor in the evaluation results.

[0031] Optionally, the calculating the case similarity corresponding to the mapped feature case includes:

[0032] Calculate the case similarity corresponding to the mapped feature case.

[0033] Optionally, combining the quantitative relationship with the oral condition of the patient to be tested to construct an oral neoplasia assessment system corresponding to the patient to be tested includes:

[0034] Analyzing the potential relationship between the quantitative relationship and the oral state corresponding to the patient to be tested;

[0035] Determining a tumorigenesis assessment index corresponding to the potential relationship;

[0036] Based on the neoplasia assessment index, setting the neoplasia index weight corresponding to the patient to be detected;

[0037] Based on the neoplasia index weights, an oral neoplasia evaluation system corresponding to the patient to be tested is constructed.

[0038] Optionally, determining the suspected tumor area corresponding to the oral cavity of the patient to be tested based on the oral neoplasia assessment system includes:

[0039] Summarizing various data in the oral neoplasia evaluation system to obtain integrated evaluation data;

[0040] Based on the integrated evaluation data, an image of the oral cavity of the patient to be tested is collected to obtain an oral cavity collection image;

[0041] Comparing the oral cavity collected image with a preset tumor characteristic map in detail to obtain a comparison abnormal point;

[0042] Based on the compared abnormal points, a suspected tumor area corresponding to the oral cavity of the patient to be tested is determined.

[0043] Optionally, the calculating, based on the development trend, a potential diffusion range corresponding to the suspected tumor area includes:

[0044] The potential diffusion range corresponding to the suspected tumor area is calculated.

[0045] Optionally, determining the pathological state corresponding to the oral cavity of the patient to be tested based on the potential diffusion range includes:

[0046] Analyzing diffusion characteristic information corresponding to the potential diffusion range;

[0047] Based on the diffusion characteristic information, query a preset medical case database to obtain diffusion-similar cases;

[0048] Compare the diffusion similar cases with the current state of the patient to be detected, and extract the corresponding common features;

[0049] Based on the common features, combined with preset professional pathological knowledge, a preliminary judgment is made to form a pathological hypothesis corresponding to the oral cavity of the patient to be tested;

[0050] Based on the pathological hypothesis, a pathological state corresponding to the oral cavity of the patient to be tested is determined.

[0051] Optionally, in order to solve the above problem, the present invention further provides an oral tumor detection system based on intelligence, the system comprising:

[0052] A signal calibration module is used to collect basic oral information of the patient to be detected, perform a layered scan on the oral cavity of the patient to be detected based on the basic oral information to obtain layered scan data, detect an oral cavity scan signal corresponding to the layered scan data, perform signal denoising on the oral cavity scan signal to obtain a denoised oral cavity signal, and perform calibration processing on the denoised oral cavity signal to obtain an oral cavity calibration signal;

[0053] a change pattern analysis module, for generating an oral feature set corresponding to the patient's oral cavity based on the oral calibration signal, screening core parameters indicating tumors according to the oral feature set combined with a preset large-scale oral tumor database, and analyzing the oral change pattern corresponding to the oral cavity of the patient to be tested based on the core parameters;

[0054] A system evaluation module is used to mine the degree of oral neoplasia under the oral change mode, analyze the quantitative relationship corresponding to the degree of oral neoplasia, combine the quantitative relationship with the oral state of the patient to be tested, and construct an oral neoplasia evaluation system corresponding to the patient to be tested;

[0055] a diffusion range calculation module, for determining the suspected tumor area corresponding to the oral cavity of the patient to be tested based on the oral neoplasia assessment system, identifying the regional features corresponding to the suspected tumor area, analyzing the development trend of the suspected tumor area based on the regional features, and calculating the potential diffusion range corresponding to the suspected tumor area based on the development trend;

[0056] A report generation module is used to determine the pathological state corresponding to the oral cavity of the patient to be detected based on the potential spread range, analyze the state feature cluster corresponding to the pathological state, judge the oral tumor risk level corresponding to the patient to be detected based on the state feature cluster, extract the key pathogenic factors in the oral tumor risk level, and generate an oral tumor detection report for the patient to be detected based on the key pathogenic factors.

[0057] First, the present invention provides an important basis and reference for subsequent oral tumor detection by collecting basic oral information of the patient to be detected, making the detection process more targeted and accurate, and helping to have a comprehensive understanding of the patient's oral condition, so as to more accurately judge the possibility and development trend of oral tumors. At the same time, the present invention generates an oral feature set corresponding to the patient's oral cavity based on the oral calibration signal, which can comprehensively and accurately describe the various characteristics of the patient's oral cavity, including information on oral tissue structure, morphology, physiological function, etc., which helps to more deeply understand the patient's oral health status and discover potential problems and abnormalities. The present invention can clearly understand the tumor by mining the degree of oral tumorization under the oral change pattern and analyzing the quantitative relationship corresponding to the degree of oral tumorization. The development stage and severity of the disease can help predict the development trend of the disease, so as to formulate a more targeted and effective treatment plan. The present invention determines the suspected tumor area corresponding to the oral cavity of the patient to be tested based on the oral neoplasia evaluation system, which can help to more accurately locate the focus of attention and concentrate subsequent diagnosis and treatment resources on these suspected areas, greatly increasing the chances of successful treatment and control of the disease, thereby improving the diagnosis and treatment efficiency. Furthermore, the present invention determines the pathological state corresponding to the oral cavity of the patient to be tested based on the potential spread range, which can clearly understand the degree of tumor invasion and possible harm, and can formulate a more targeted and effective treatment strategy, and can timely discover potential complication risks, take preventive measures in advance, and reduce the incidence of complications in patients during treatment. Therefore, the intelligent oral tumor detection method and system proposed by the present invention can improve the accuracy of oral tumor detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 A schematic diagram of a flow chart of an intelligent oral tumor detection method provided by an embodiment of the present invention;

[0059] Figure 2 A schematic diagram of a module of an intelligent oral tumor detection system provided by an embodiment of the present invention.

[0060] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0061] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0062] The embodiment of the present application provides an oral tumor detection method based on intelligence. The execution subject of the oral tumor detection method based on intelligence includes but is not limited to at least one of the electronic devices such as a server, a terminal, etc. that can be configured to execute the method provided in the embodiment of the present application. In other words, the oral tumor detection method based on intelligence can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc. The server can be an independent server, or it can be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN), and big data and artificial intelligence platforms.

[0063] Reference Figure 1 FIG. 1 is a flow chart of an intelligent oral tumor detection method according to an embodiment of the present invention. In this embodiment, the intelligent oral tumor detection method includes:

[0064] S1. Collect basic oral information of the patient to be tested, perform a layered scan on the oral cavity of the patient to be tested based on the basic oral information to obtain layered scan data, detect the oral scan signal corresponding to the layered scan data, perform signal denoising on the oral scan signal to obtain a denoised oral signal, and perform calibration processing on the denoised oral signal to obtain an oral calibration signal.

[0065] The present invention provides an important basis and reference for subsequent oral tumor detection by collecting basic oral information of the patient to be detected, making the detection process more targeted and accurate, and helping to have a comprehensive understanding of the patient's oral condition, so as to more accurately judge the possibility and development trend of oral tumors.

[0066] Among them, the basic oral information refers to the oral medical history, family medical history, living habits and other related information of the patient to be tested. Optionally, the collection of the basic oral information of the patient to be tested can be achieved by querying medical records, such as: querying the patient's previous oral examination report, diagnosis results, etc. through the hospital's electronic medical record system.

[0067] Based on the basic oral information, the present invention performs a layered scan on the oral cavity of the patient to be tested to obtain layered scanning data, which can provide a more comprehensive and detailed understanding of the structure and tissue conditions inside the patient's oral cavity, and can accurately observe the details of each layer of the oral cavity, including mucosa, teeth, alveolar bone, etc., which is helpful to discover potential abnormal lesions.

[0068] The layered scanning data refers to the detailed data obtained after layered scanning of the oral cavity according to the layered planning range, and these data can more accurately display the structure and characteristics of each layer inside the oral cavity.

[0069] As an embodiment of the present invention, based on the oral basic information, a layered scan is performed on the oral cavity of the patient to be detected to obtain layered scan data, including: querying basic parameters corresponding to the oral basic information; based on the basic parameters, performing a preliminary scan on the oral cavity of the patient to be detected to obtain initial scan oral data; denoising the initial scan oral data to obtain denoised initial scan oral data; performing layered planning on the denoised initial scan oral data to obtain a layered planning range; based on the layered planning range, performing a layered scan on the oral cavity of the patient to be detected to obtain layered scan data.

[0070] Among them, the basic parameters refer to various parameters related to the basic information of the oral cavity, such as parameters related to the size, shape, tissue structure and other characteristics of the oral cavity; the initial scan oral data refers to the data obtained by performing a preliminary scan of the oral cavity of the patient to be tested, and these data preliminarily reflect the overall situation of the oral cavity; the denoised initial scan oral data refers to the data obtained after denoising the initial scan oral data, removing noise interference and more accurately reflecting the actual situation of the oral cavity; the hierarchical planning range refers to the specific range and hierarchical division of the hierarchical scanning of the oral cavity determined based on the denoised initial scan oral data.

[0071] Furthermore, the query of the basic parameters corresponding to the basic oral information can be achieved through an artificial intelligence algorithm, such as: using a neural network algorithm to learn historical data to infer the patient's basic oral parameters; the preliminary scanning of the oral cavity of the patient to be tested can be achieved through an oral CT scanning method, such as: using an oral CT device to scan the patient's oral cavity to obtain the three-dimensional structural information of the oral cavity; the denoising of the initial scan oral data can be achieved through a denoising algorithm, such as: median filtering, wavelet transform and other algorithms; the hierarchical planning of the denoised initial scan oral data can be achieved through a deep learning algorithm, such as: using a semantic segmentation model in deep learning to perform hierarchical planning on oral data; the hierarchical scanning of the oral cavity of the patient to be tested can be achieved through a multi-layer spiral CT scan, such as: using a multi-layer spiral CT device to perform precise scanning according to the hierarchical planning range.

[0072] By detecting the oral scanning signal corresponding to the layered scanning data, the present invention can gain a deeper understanding of the physiological and pathological conditions inside the oral cavity, and can provide information on the structure, density, blood flow, etc. of the oral tissue, which is helpful to discover potential abnormalities, such as early signs of tumors.

[0073] Among them, the oral scanning signal refers to a signal reflecting the internal condition of the oral cavity obtained by detecting and analyzing the layered scanning data, which may include information such as the physical properties and physiological activities of the oral tissue. Optionally, the detection of the oral scanning signal corresponding to the layered scanning data can be achieved through imaging technology methods, such as: magnetic resonance imaging, ultrasonic imaging and other methods.

[0074] Furthermore, the present invention performs signal denoising on the oral scanning signal to obtain a denoised oral signal, which can more accurately reflect the actual situation inside the oral cavity, remove the interference of noise, and can more clearly observe and analyze the characteristics of oral tissue, which helps to improve the accuracy of diagnosis and make subsequent analysis and processing more precise and effective.

[0075] Among them, the denoised oral signal refers to the signal obtained after denoising the oral scanning signal, which removes the mixed noise components in the signal, so that the signal can more accurately reflect the real information inside the oral cavity and reduce interference and errors. Optionally, the signal denoising of the oral scanning signal can be achieved through a signal denoising method, such as: wavelet transform, median filtering, adaptive filtering and other methods.

[0076] Furthermore, the present invention obtains an oral calibration signal by calibrating the denoised oral signal, which can eliminate deviations and errors in the signal, improve the accuracy of subsequent analysis and diagnosis, provide a more accurate reference basis, and thus accurately judge the health status of the oral cavity. It can more clearly display the characteristics and changes of oral tissues, and help to discover potential problems and abnormalities.

[0077] Among them, the oral calibration signal refers to the signal obtained after the denoised oral signal is calibrated, which has higher accuracy and reliability and can more accurately reflect the actual situation inside the oral cavity. Optionally, the calibration processing of the denoised oral signal can be achieved through a statistical analysis calibration method, such as: by performing statistical analysis on a large amount of oral signal data, determining the normal range and characteristics of the signal, and thus calibrating the denoised oral signal.

[0078] S2. Generate an oral feature set corresponding to the patient's oral cavity based on the oral calibration signal, screen core parameters indicating tumors according to the oral feature set in combination with a preset large-scale oral tumor database, and analyze the oral change pattern corresponding to the oral cavity of the patient to be tested based on the core parameters.

[0079] The present invention generates an oral feature set corresponding to the patient's oral cavity based on the oral calibration signal, which can comprehensively and accurately describe various characteristics of the patient's oral cavity, including information on oral tissue structure, morphology, physiological function, etc., which helps to more deeply understand the patient's oral health status and discover potential problems and abnormalities.

[0080] Among them, the oral feature set refers to a set generated based on core features, which comprehensively and accurately describes various characteristics of the patient's oral cavity, including structure, function, pathology and other aspects, providing an important basis for subsequent oral tumor detection.

[0081] As an embodiment of the present invention, generating an oral feature set corresponding to the patient's oral cavity based on the oral calibration signal includes: performing an in-depth analysis on the oral calibration signal and extracting key feature parameters in the oral calibration signal; defining a feature identifier corresponding to the oral calibration signal based on the key feature parameters; performing association analysis on the feature identifier to obtain a feature association matrix; screening core features in the feature association matrix; and generating an oral feature set corresponding to the patient's oral cavity based on the core features.

[0082] Among them, the key feature parameters refer to parameters with important significance and representativeness extracted from the oral calibration signal, which can reflect the characteristics of the structure, function or pathological state of the oral cavity; the feature identifier refers to a symbol or code defined based on the key feature parameters for identifying and distinguishing different features; the feature association matrix refers to a matrix obtained after association analysis of the feature identifiers, which shows the degree of association and mutual relationship between different features.

[0083] Furthermore, the in-depth analysis of the oral calibration signal can be achieved through signal analysis methods, such as spectrum analysis, waveform analysis and other methods; the extraction of key feature parameters in the oral calibration signal can be achieved through feature selection algorithms, such as filtering feature selection, wrapping feature selection or embedded feature selection, to screen out the most representative and distinguishing key feature parameters; the definition of the feature identifier corresponding to the oral calibration signal can be achieved through a hash function, such as using a hash function to calculate the feature parameters to obtain a hash value as a feature identifier; the association analysis of the feature identifier can be achieved through an association rule mining algorithm, such as an Apriori algorithm or an FP-Growth algorithm; the screening of core features in the feature association matrix can be achieved through a machine learning model, such as a support vector machine or a neural network, to train a classification or regression model, and then screen out the core features according to the weights or coefficients of the features in the model; the generation of the oral feature set corresponding to the patient's mouth can be achieved through set operations, such as using set operations, such as union, intersection or difference, to combine the screened core features into an oral feature set.

[0084] The present invention combines the oral feature set with a preset large-scale oral tumor database to screen the core parameters indicating tumors, which helps to improve the accuracy and specificity of oral tumor detection. It can more accurately identify tumor-related features and parameters, and can detect potential tumor lesions early. Signs of tumors can be detected at an early stage, so that treatment measures can be taken in time.

[0085] Among them, the preset large-scale oral tumor database refers to a database established in advance that contains the oral characteristics and related diagnostic information of a large number of oral tumor patients; the tumor indication refers to the characteristics or indicators that can indicate the possible presence of a tumor in the oral cavity; the core parameters refer to the important parameters determined in the evaluation results to play a key role in tumor indication.

[0086] As an embodiment of the present invention, the method of screening core parameters indicating tumors based on the oral feature set in combination with a preset large-scale oral tumor database includes: performing feature mapping between the features in the oral feature set and the features in the preset large-scale oral tumor database to obtain mapped feature cases; calculating the case similarity corresponding to the mapped feature cases; determining the corresponding similar case set based on the case similarity; performing frequency statistics on the features in the similar case set to obtain the feature occurrence frequency; screening high-frequency features in the feature occurrence frequency; performing correlation analysis on the high-frequency features to obtain a feature correlation matrix; extracting strongly correlated feature combinations in the feature correlation matrix; performing clinical evaluation on the strongly correlated feature combinations to obtain evaluation results; and determining core parameters indicating tumors in the evaluation results.

[0087] Among them, the mapped feature cases refer to related cases obtained by matching the features in the oral feature set with the features in the large-scale oral tumor database; the case similarity refers to the quantitative expression of the degree of similarity in features between the current oral feature set to be detected and the existing cases in the database; the similar case set refers to a set of a series of cases similar to the current features to be detected, which are determined based on the calculated case similarity; the feature occurrence frequency refers to the number of times each feature in the similar case set appears in these cases; the high-frequency feature refers to the feature that appears more frequently in the feature occurrence frequency; the feature correlation matrix refers to a matrix used to represent the degree of mutual relationship between high-frequency features; the strongly correlated feature combination refers to a combination of features with strong correlation in the feature correlation matrix; the evaluation result refers to the conclusion on its relationship with tumor indication obtained after clinical evaluation of the strongly correlated feature combination.

[0088] Furthermore, the feature mapping of the features in the oral feature set with the features in a preset large-scale oral tumor database can be achieved through hash mapping, such as: fast mapping of features through a consistent hashing algorithm; the calculation of the case similarity corresponding to the mapped feature case can be achieved through the following calculation method; the determination of the corresponding similar case set can be achieved through a sorting and screening method, such as: arranging the cases in descending order according to the similarity, and selecting the first several cases as the similar case set; the frequency statistics of the features in the similar case set can be achieved through a statistical query tool, such as: SQL COUNT function for statistics; the screening of high-frequency features in the frequency of occurrence of the features can be achieved by setting a frequency threshold method, For example, features with an occurrence frequency higher than a specific threshold are determined as high-frequency features; the correlation analysis of the high-frequency features can be achieved through correlation analysis methods, such as Pearson correlation coefficient, Spearman rank correlation coefficient, etc.; the extraction of strongly correlated feature combinations in the feature correlation matrix can be achieved through clustering analysis methods, such as clustering features with higher correlation into one category, and extracting strongly correlated feature combinations from them; the clinical evaluation of the strongly correlated feature combinations can be achieved through clinical indicator control methods, such as comparing and evaluating the feature combinations with known clinical diagnostic indicators; the determination of the core parameters indicating tumors in the evaluation results can be achieved through multiple rounds of evaluation and screening methods, such as performing multiple evaluations and screenings to gradually determine the core parameters.

[0089] As an embodiment of the present invention, the calculating of the case similarity corresponding to the mapped feature case includes:

[0090] The case similarity corresponding to the mapped feature case is calculated using the following formula:

[0091]

[0092] Among them, BSD represents the case similarity corresponding to the mapped characteristic case, n represents the number of feature vectors in the mapped characteristic case, i represents the number index of feature vectors in the mapped characteristic case, and a i represents the ith eigenvector value of case a in the mapped feature case, b i Represents the i-th eigenvector value of case b in the mapped feature cases.

[0093] Furthermore, the present invention analyzes the oral change pattern corresponding to the oral cavity of the patient to be tested based on the core parameters, which can more accurately determine the development trend of the disease, help to adjust the treatment plan in time, and can discover potential problems and early signs of abnormal changes. Even before symptoms are obvious, subtle but critical changes can be captured, thereby achieving early intervention and treatment.

[0094] Among them, the oral change pattern refers to the changes with certain rules and characteristics presented by the core parameters of the oral cavity of the patient to be tested within a period of time, including but not limited to the increase or decrease trend, fluctuation amplitude, periodic changes, etc. of the core parameters. Optionally, the analysis of the oral change pattern corresponding to the oral cavity of the patient to be tested can be achieved through a time series analysis method, such as: using an autoregressive moving average model or a differential integrated moving average autoregressive model to predict and analyze the future trend of the parameters.

[0095] S3. Mining the degree of oral neoplasia under the oral change pattern, analyzing the quantitative relationship corresponding to the degree of oral neoplasia, combining the quantitative relationship with the oral condition corresponding to the patient to be tested, and constructing an oral neoplasia evaluation system corresponding to the patient to be tested.

[0096] By mining the degree of oral neoplasia under the oral change pattern and analyzing the quantitative relationship corresponding to the degree of oral neoplasia, the present invention can clearly understand the development stage and severity of the tumor, help predict the development trend of the disease, and thus formulate a more targeted and effective treatment plan.

[0097] Among them, the degree of oral neoplasia refers to the degree of transformation of internal oral tissues into tumors, including a comprehensive evaluation of the number, size, distribution range of tumor cells, and the degree of invasion of surrounding tissues; the quantitative relationship refers to the association between the degree of oral neoplasia and various related factors (such as specific core parameters, time, treatment measures, etc.), which can be described by specific numerical values ​​and mathematical formulas. Optionally, the mining of the degree of oral neoplasia under the oral change pattern can be achieved through a biomarker detection method, such as: detecting the level of a specific biomarker to indirectly reflect the degree of neoplasia; the analysis of the quantitative relationship corresponding to the degree of oral neoplasia can be achieved through the TensorFlow framework, such as: building a deep neural network through the TensorFlow framework, and training a large amount of oral neoplasia data to predict the degree of neoplasia.

[0098] Furthermore, the present invention constructs an oral neoplasia assessment system corresponding to the patient to be tested by combining the quantitative relationship with the oral condition of the patient to be tested.

[0099] Among them, the oral state refers to various physiological and pathological manifestations inside the oral cavity of the patient to be tested, including but not limited to the color, texture, integrity of the oral mucosa, the health status, arrangement and occlusion of the teeth, saliva secretion, the presence of ulcers, lumps, inflammation, bleeding and other abnormalities, as well as the morphology of the tongue, the characteristics of the tongue coating and other comprehensive descriptions of the state; the oral neoplasia assessment system refers to a complete and systematic framework, which includes a series of interrelated assessment methods, standards, indicators and their weights and other elements, aiming to accurately, comprehensively and scientifically assess and judge the oral neoplasia status of the patient to be tested.

[0100] As an embodiment of the present invention, the quantitative relationship is combined with the oral condition corresponding to the patient to be detected to construct an oral neoplasia evaluation system corresponding to the patient to be detected, including: analyzing the potential relationship between the quantitative relationship and the oral condition corresponding to the patient to be detected; determining the neoplasia evaluation index corresponding to the potential relationship; based on the neoplasia evaluation index, setting the neoplasia index weight corresponding to the patient to be detected; based on the neoplasia index weight, constructing the oral neoplasia evaluation system corresponding to the patient to be detected.

[0101] Among them, the potential relationship refers to the mutual correlation and influence mode that may exist but has not yet been clearly recognized or clearly expressed between the various data covered by the quantitative relationship and the specific oral status manifestations of the patient to be tested; the neoplasia assessment index refers to those specific parameters or characteristics that can specifically quantify and reflect the degree, nature, progression, etc. of oral neoplasia in the patient to be tested; the neoplasia index weight refers to the relative numerical ratio assigned based on the importance of the neoplasia assessment index for evaluating the oral neoplasia situation of the patient to be tested.

[0102] Furthermore, the analysis of the potential relationship between the quantitative relationship and the oral state corresponding to the patient to be tested can be achieved through a correlation analysis method, such as: using the corr function in Python's Pandas library to calculate the Pearson correlation; the determination of the tumor assessment index corresponding to the potential relationship can be achieved through a decision tree analysis method, such as: by constructing a decision tree to screen out indicators that are important for tumor assessment; the setting of the tumor index weight corresponding to the patient to be tested can be achieved through an entropy weight method, such as: determining the weight according to the degree of variation of the indicator; the construction of the oral tumor assessment system corresponding to the patient to be tested can be achieved through a neural network model, such as: using TensorFlow to build a neural network model to construct an evaluation system.

[0103] S4. Based on the oral neoplasia assessment system, determine the suspected tumor area corresponding to the oral cavity of the patient to be tested, identify the regional characteristics corresponding to the suspected tumor area, analyze the development trend of the suspected tumor area based on the regional characteristics, and calculate the potential diffusion range corresponding to the suspected tumor area based on the development trend.

[0104] The present invention is based on the oral neoplasia assessment system to determine the suspected tumor area corresponding to the oral cavity of the patient to be tested, which can help to more accurately locate the focus of attention and concentrate subsequent diagnosis and treatment resources on these suspected areas, greatly increasing the chances of successful treatment and control of the disease, thereby improving the efficiency of diagnosis and treatment.

[0105] The suspected tumor area refers to the area in the oral cavity of the patient to be tested where the tumor may exist, which is inferred based on the comparison of abnormal points.

[0106] As an embodiment of the present invention, the method of determining the suspected tumor area corresponding to the oral cavity of the patient to be detected based on the oral neoplasia evaluation system includes: summarizing various data in the oral neoplasia evaluation system to obtain integrated evaluation data; based on the integrated evaluation data, performing image acquisition on the oral cavity of the patient to be detected to obtain an oral acquisition image; performing a detailed comparison of the oral acquisition image with a preset tumor feature map to obtain comparison abnormal points; and determining the suspected tumor area corresponding to the oral cavity of the patient to be detected based on the comparison abnormal points.

[0107] Among them, the integrated evaluation data refers to the comprehensive data obtained by summarizing and organizing the various quantitative relationships, evaluation indicators and their weights in the oral neoplasia evaluation system; the oral acquisition image refers to the image information obtained by photographing or scanning the oral cavity of the patient to be tested through specific equipment and technology; the preset tumor feature map refers to a pre-established detailed image set containing various known tumors that may present characteristics in the oral cavity, such as shape, color, texture, etc.; the contrast abnormal point refers to the part or feature point that does not conform to the normal characteristics and has differences found in the process of comparing the oral acquisition image with the preset tumor feature map.

[0108] Furthermore, the summary of various data in the oral neoplasia assessment system can be achieved through database management tools, such as MySQL, Oracle and other tools; the image acquisition of the oral cavity of the patient to be detected can be achieved through an oral endoscope, such as using a professional oral endoscope device to directly obtain images of the inside of the oral cavity; the detailed comparison of the oral captured image with the preset tumor feature map can be achieved through an image matching algorithm, such as using the SIFT algorithm in the OpenCV library to extract and match image features; the determination of the suspected tumor area corresponding to the oral cavity of the patient to be detected can be achieved through an image segmentation algorithm, such as threshold-based segmentation, region growing algorithm and other algorithms.

[0109] By identifying the regional features corresponding to the suspected tumor area, the present invention can provide a key basis for subsequent accurate diagnosis, and can more accurately determine whether the suspected area is really a tumor, avoiding misdiagnosis and missed diagnosis, so that the most suitable treatment method for the patient can be selected in a targeted manner.

[0110] Among them, the regional characteristics refer to various characteristics presented by the suspected tumor area, including but not limited to characteristics in terms of shape, size, color, texture, boundary clarity, internal structure, and relationship with surrounding tissues. Optionally, the identification of the regional characteristics corresponding to the suspected tumor area can be achieved through ImageJ tools, such as: using ImageJ to measure the area, perimeter, etc. of the suspected area.

[0111] Furthermore, the present invention analyzes the development trend of the suspected tumor area based on the regional characteristics, and can predict the growth rate, invasion range and possible degree of malignancy of the tumor according to the development trend, so as to plan a more accurate and effective treatment plan in advance and avoid worsening of the disease due to delayed treatment.

[0112] Among them, the development trend refers to the direction and degree of change that the suspected tumor area may show over a period of time, including changes in the size, shape, texture, growth rate, whether the tumor spreads, etc. Optionally, the analysis of the development trend of the suspected tumor area can be achieved through a time series analysis method, such as: observing the numerical changes of regional characteristics at different time points to predict future trends.

[0113] Furthermore, based on the development trend, the present invention calculates the potential diffusion range corresponding to the suspected tumor area, which helps to determine the scope of subsequent surgical resection, ensure the complete removal of tumor cells, and reduce the risk of recurrence. The irradiation area of ​​radiotherapy and the range of action of chemotherapy drugs can be accurately located according to the diffusion range, thereby improving the treatment effect while reducing damage to normal tissues.

[0114] The potential spread range refers to the area where tumor cells may spread and invade in the future.

[0115] As an embodiment of the present invention, the calculating, based on the development trend, the potential diffusion range corresponding to the suspected tumor area includes:

[0116] The potential diffusion range corresponding to the suspected tumor area is calculated using the following formula:

[0117]

[0118] Wherein, QF represents the potential diffusion range corresponding to the suspected tumor area, q represents the number of tumor development features in the suspected tumor area, p represents the number index of tumor development features in the suspected tumor area, and t 1 and t 2 Respectively represent the starting time point and the ending time point of observing the suspected tumor area, α p represents the weight of the p-th tumor development feature, F p (t) represents the function of the development characteristics of the pth tumor over time t, F p (t) / dt represents the rate of change of the pth tumor development characteristic, m represents the number of external factors corresponding to the development trend, j represents the number index of external factors corresponding to the development trend, β j represents the weight of the jth external factor, H j Represents the quantitative value of the jth external factor.

[0119] In detail, the weight of the tumor development characteristic refers to a relative numerical value used to measure the influence of each tumor development characteristic (such as the growth rate of tumor size, the degree of morphological change, etc.) on the potential spread range; the change function refers to a mathematical expression that describes the change pattern of tumor development characteristics (such as tumor volume, cell density, etc.) over time; the change rate of the tumor development characteristic refers to the change amount of the tumor development characteristic per unit time, reflecting the speed of tumor development; the quantitative value of the external factor refers to the numerical representation of other external factors that may affect the spread range of the tumor (such as the patient's immune system status, living habits, treatment intervention effects, etc.).

[0120] S5. Based on the potential spread range, determine the pathological state corresponding to the oral cavity of the patient to be detected, analyze the state feature cluster corresponding to the pathological state, judge the oral tumor risk level corresponding to the patient to be detected based on the state feature cluster, extract the key pathogenic factors in the oral tumor risk level, and generate an oral tumor detection report for the patient to be detected based on the key pathogenic factors.

[0121] The present invention determines the pathological state of the oral cavity of the patient to be tested based on the potential spread range, can clearly understand the degree of tumor invasion and the possible harm caused, can formulate more targeted and effective treatment strategies, can timely discover potential complication risks, take preventive measures in advance, and reduce the incidence of complications in patients during treatment.

[0122] The pathological state refers to a clear diagnosis result of the comprehensive conditions such as the nature, type, severity and development stage of the oral disease of the patient to be tested.

[0123] As an embodiment of the present invention, determining the pathological state corresponding to the oral cavity of the patient to be detected based on the potential diffusion range includes: analyzing diffusion feature information corresponding to the potential diffusion range; querying a preset medical case database based on the diffusion feature information to obtain diffusion similar cases; comparing the diffusion similar cases with the current state of the patient to be detected, and extracting their corresponding common features; making a preliminary judgment based on the common features and combining preset professional pathological knowledge to form a pathological hypothesis corresponding to the oral cavity of the patient to be detected; and determining the pathological state corresponding to the oral cavity of the patient to be detected based on the pathological hypothesis.

[0124] Among them, the diffusion characteristic information refers to the specific characteristics of the potential diffusion range, such as a detailed description of the diffusion direction, speed, range and involved tissue types; the preset medical case database refers to a collection of relevant information of a large number of known cases that has been established and stored in advance, including data such as the patient's condition, diagnosis results, treatment plan and prognosis; the diffusion similar cases refer to past cases in the medical case database that have similar characteristics to the potential diffusion range of the current patient to be tested; the common characteristics refer to the same or similar characteristics shared by the diffusion similar cases and the patients to be tested in terms of disease manifestations, diffusion characteristics, etc.; the preset professional pathology knowledge refers to a systematic knowledge system of pathology theories, principles, diagnostic standards and treatment experience that is pre-set and accumulated; the pathology hypothesis refers to a preliminary speculation and assumption about the oral pathology of the patient to be tested based on common characteristics and professional pathology knowledge.

[0125] Furthermore, the analysis of the diffusion characteristic information corresponding to the potential diffusion range can be achieved through data visualization analysis, such as: Tableau, PowerBI and other tools; the query of the preset medical case database can be achieved through SQL tools, such as: using SQL statements in a relational database for precise and fuzzy queries; the comparison of the diffusion similar cases with the current status of the patient to be detected can be achieved through a text similarity calculation method, such as: using the sklearn library in Python to calculate text cosine similarity; the preliminary judgment based on preset professional pathological knowledge can be achieved through a rule-based reasoning system, such as: using an expert system tool, inputting patient characteristics and rules, and making a preliminary judgment; the determination of the pathological state corresponding to the oral cavity of the patient to be detected can be achieved through MATLAB tools, such as: using the data fusion function in MATLAB to integrate and analyze multiple data.

[0126] Furthermore, the present invention helps to have a more comprehensive and in-depth understanding of the patient's condition by analyzing the state feature clusters corresponding to the pathological state, and the analysis of the state feature clusters can clearly grasp various aspects of the pathological state, including the manifestation, severity, development stage, etc. of symptoms, thereby providing a detailed basis for precise treatment.

[0127] Among them, the state feature cluster refers to a collection of a series of features associated with a specific pathological state. These features are interrelated and influence each other, and together describe the overall picture of the pathological state. Optionally, the analysis of the state feature cluster corresponding to the pathological state can be achieved through a clustering analysis method, such as: applying the K-Means algorithm in Python's Scikit-learn library to cluster pathological features.

[0128] Furthermore, the present invention determines the oral tumor risk level corresponding to the patient to be detected based on the state feature cluster, extracts the key pathogenic factors in the oral tumor risk level, and clarifies the oral tumor risk level to have a clear and quantitative understanding of the patient's condition, thereby rationally allocating medical resources and providing more timely and comprehensive treatment interventions for high-risk patients.

[0129] Among them, the oral tumor risk level refers to different risk levels divided after comprehensive evaluation of various status feature clusters of the patient, such as low risk, medium risk and high risk, which are used to indicate the possibility of the patient suffering from oral tumors and the severity of the disease; the key pathogenic factors refer to factors that play an important role and have a greater influence in the occurrence and development of oral tumors. Optionally, the determination of the oral tumor risk level corresponding to the patient to be tested can be achieved through a risk assessment model, such as: using the glm function in the R language to construct a logistic regression model for risk assessment; the extraction of the key pathogenic factors in the oral tumor risk level can be achieved through a feature importance assessment method, such as: in Python's Scikit-learn library, using a random forest classifier to calculate the importance score of the feature.

[0130] Furthermore, the present invention generates an oral tumor detection report for the patient to be detected based on the key pathogenic factors, providing the patient with a clear, comprehensive and personalized description of the condition, enabling the patient to intuitively understand the key causes of his or her illness, thereby better understanding the condition and increasing the degree of attention and cooperation to treatment.

[0131] Among them, the oral tumor detection report refers to a written document that integrates the various examination results of the patient to be tested, key pathogenic factor analysis, risk assessment, diagnostic conclusions and treatment recommendations, etc., and is intended to clearly and accurately convey relevant information about the patient's oral tumor to the patient, doctor and other relevant personnel. Optionally, the generation of the oral tumor detection report for the patient to be tested can be achieved through data visualization tools, such as: Tableau, PowerBI and other tools.

[0132] First, the present invention provides an important basis and reference for subsequent oral tumor detection by collecting basic oral information of the patient to be detected, making the detection process more targeted and accurate, and helping to have a comprehensive understanding of the patient's oral condition, so as to more accurately judge the possibility and development trend of oral tumors. At the same time, the present invention generates an oral feature set corresponding to the patient's oral cavity based on the oral calibration signal, which can comprehensively and accurately describe the various characteristics of the patient's oral cavity, including information on oral tissue structure, morphology, physiological function, etc., which helps to more deeply understand the patient's oral health status and discover potential problems and abnormalities. The present invention can clearly understand the tumor by mining the degree of oral tumorization under the oral change pattern and analyzing the quantitative relationship corresponding to the degree of oral tumorization. The development stage and severity of the disease can help predict the development trend of the disease, so as to formulate a more targeted and effective treatment plan. The present invention determines the suspected tumor area corresponding to the oral cavity of the patient to be tested based on the oral neoplasia evaluation system, which can help to more accurately locate the focus of attention and concentrate subsequent diagnosis and treatment resources on these suspected areas, greatly increasing the chances of successful treatment and control of the disease, thereby improving the diagnosis and treatment efficiency. Furthermore, the present invention determines the pathological state corresponding to the oral cavity of the patient to be tested based on the potential spread range, which can clearly understand the degree of tumor invasion and possible harm, and can formulate a more targeted and effective treatment strategy, and can timely discover potential complication risks, take preventive measures in advance, and reduce the incidence of complications in patients during treatment. Therefore, the intelligent oral tumor detection method and system proposed by the present invention can improve the accuracy of oral tumor detection.

[0133] like Figure 2 FIG. 1 is a functional module diagram of an intelligent oral tumor detection system provided in accordance with an embodiment of the present invention.

[0134] The intelligent oral tumor detection system 200 of the present invention can be installed in an electronic device. According to the functions to be implemented, the intelligent oral tumor detection system 200 can include a signal calibration module 201, a change pattern analysis module 202, a system evaluation module 203, a diffusion range calculation module 204 and a report generation module 205. The module of the present invention can also be called a unit, which refers to a series of computer program segments that can be executed by an electronic device processor and can complete fixed functions, which are stored in the memory of the electronic device.

[0135] In this embodiment, the functions of each module / unit are as follows:

[0136] The signal calibration module 201 is used to collect basic oral information of the patient to be detected, perform layered scanning on the oral cavity of the patient to be detected based on the basic oral information to obtain layered scanning data, detect the oral cavity scanning signal corresponding to the layered scanning data, perform signal denoising on the oral cavity scanning signal to obtain a denoised oral cavity signal, and perform calibration processing on the denoised oral cavity signal to obtain an oral cavity calibration signal;

[0137] The change pattern analysis module 202 is used to generate an oral feature set corresponding to the patient's oral cavity based on the oral calibration signal, screen core parameters indicating tumors according to the oral feature set combined with a preset large-scale oral tumor database, and analyze the oral change pattern corresponding to the oral cavity of the patient to be tested based on the core parameters;

[0138] The system evaluation module 203 is used to mine the degree of oral neoplasia under the oral change mode, analyze the quantitative relationship corresponding to the degree of oral neoplasia, combine the quantitative relationship with the oral state of the patient to be tested, and construct an oral neoplasia evaluation system corresponding to the patient to be tested;

[0139] The diffusion range calculation module 204 is used to determine the suspected tumor area corresponding to the oral cavity of the patient to be tested based on the oral neoplasia assessment system, identify the regional characteristics corresponding to the suspected tumor area, analyze the development trend of the suspected tumor area based on the regional characteristics, and calculate the potential diffusion range corresponding to the suspected tumor area based on the development trend;

[0140] The report generation module 205 is used to determine the pathological state corresponding to the oral cavity of the patient to be detected based on the potential spread range, analyze the state feature cluster corresponding to the pathological state, judge the oral tumor risk level corresponding to the patient to be detected based on the state feature cluster, extract the key pathogenic factors in the oral tumor risk level, and generate an oral tumor detection report for the patient to be detected based on the key pathogenic factors.

[0141] In detail, each module described in the intelligent oral tumor detection system 200 described in the embodiment of the present invention adopts the same technical means as the intelligent oral tumor detection method described in the accompanying drawings when used, and can produce the same technical effects, which will not be repeated here.

[0142] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. An oral tumor detection method based on intelligence, characterized in that: The method comprises: Collecting basic oral information of the patient to be tested, performing layered scanning on the oral cavity of the patient to be tested based on the basic oral information to obtain layered scanning data, detecting an oral scanning signal corresponding to the layered scanning data, performing signal denoising on the oral scanning signal to obtain a denoised oral signal, and performing calibration processing on the denoised oral signal to obtain an oral calibration signal; Based on the oral calibration signal, an oral feature set corresponding to the patient's oral cavity is generated; based on the oral feature set combined with a preset large-scale oral tumor database, core parameters indicating tumors are screened; and based on the core parameters, an oral change pattern corresponding to the oral cavity of the patient to be tested is analyzed; Mining the degree of oral neoplasia under the oral change pattern, analyzing the quantitative relationship corresponding to the degree of oral neoplasia, combining the quantitative relationship with the oral state of the patient to be tested, and constructing an oral neoplasia evaluation system corresponding to the patient to be tested; Based on the oral neoplasia assessment system, determining the suspected tumor area corresponding to the oral cavity of the patient to be tested, identifying the regional features corresponding to the suspected tumor area, analyzing the development trend of the suspected tumor area based on the regional features, and calculating the potential diffusion range corresponding to the suspected tumor area based on the development trend; Based on the potential spread range, the pathological state corresponding to the oral cavity of the patient to be tested is determined, the state feature cluster corresponding to the pathological state is analyzed, and based on the state feature cluster, the oral tumor risk level corresponding to the patient to be tested is judged, and the key pathogenic factors in the oral tumor risk level are extracted. Based on the key pathogenic factors, an oral tumor detection report for the patient to be tested is generated.

2. The intelligent oral tumor detection method according to claim 1, characterized in that: The step of performing a layered scan on the oral cavity of the patient to be detected based on the oral cavity basic information to obtain layered scan data includes: Query the basic parameters corresponding to the oral basic information; Based on the basic parameters, a preliminary scan is performed on the oral cavity of the patient to be tested to obtain preliminary scan oral cavity data; Performing denoising on the initial oral scan data to obtain denoised initial oral scan data; Performing hierarchical planning on the denoised initial scan oral data to obtain a hierarchical planning range; Based on the hierarchical planning range, a hierarchical scan is performed on the oral cavity of the patient to be detected to obtain hierarchical scan data.

3. The intelligent oral tumor detection method according to claim 1, characterized in that: Generating an oral feature set corresponding to the patient's oral cavity based on the oral calibration signal includes: Performing in-depth analysis on the oral calibration signal and extracting key characteristic parameters from the oral calibration signal; Based on the key characteristic parameters, defining a characteristic identifier corresponding to the oral calibration signal; Performing association analysis on the feature identifiers to obtain a feature association matrix; Screening the core features in the feature association matrix; Based on the core features, an oral feature set corresponding to the patient's oral cavity is generated.

4. The intelligent oral tumor detection method according to claim 1, characterized in that: The method of screening core parameters indicating tumors based on the oral feature set in combination with a preset large-scale oral tumor database includes: Performing feature mapping between the features in the oral feature set and the features in a preset large-scale oral tumor database to obtain mapped feature cases; Calculating case similarity corresponding to the mapped feature case; Based on the case similarity, determining the corresponding similar case set; Performing frequency statistics on the features in the similar case set to obtain the frequency of feature occurrence; Screening high-frequency features from the occurrence frequencies of the features; performing correlation analysis on the high-frequency features to obtain a feature correlation matrix; Extracting strongly correlated feature combinations in the feature correlation matrix; Performing a clinical evaluation on the strongly correlated feature combination to obtain an evaluation result; Determine the core parameters indicating the presence of a tumor in the evaluation results.

5. The intelligent oral tumor detection method according to claim 1, characterized in that: The calculating the case similarity corresponding to the mapped feature case includes: Calculate the case similarity corresponding to the mapped feature case.

6. The intelligent oral tumor detection method according to claim 1, characterized in that: The step of combining the quantitative relationship with the oral condition of the patient to be tested to construct an oral neoplasia assessment system corresponding to the patient to be tested includes: Analyzing the potential relationship between the quantitative relationship and the oral state corresponding to the patient to be tested; Determining a tumorigenesis assessment index corresponding to the potential relationship; Based on the neoplasia assessment index, setting the neoplasia index weight corresponding to the patient to be detected; Based on the neoplasia index weights, an oral neoplasia evaluation system corresponding to the patient to be tested is constructed.

7. The intelligent oral tumor detection method according to claim 1, characterized in that: Determining the suspected tumor area corresponding to the oral cavity of the patient to be tested based on the oral neoplasia assessment system includes: Summarizing various data in the oral neoplasia evaluation system to obtain integrated evaluation data; Based on the integrated evaluation data, an image of the oral cavity of the patient to be tested is collected to obtain an oral cavity collection image; Comparing the oral cavity collected image with a preset tumor characteristic map in detail to obtain a comparison abnormal point; Based on the compared abnormal points, a suspected tumor area corresponding to the oral cavity of the patient to be tested is determined.

8. The intelligent oral tumor detection method according to claim 1, characterized in that: The calculating, based on the development trend, the potential diffusion range corresponding to the suspected tumor area includes: The potential diffusion range corresponding to the suspected tumor area is calculated.

9. The intelligent oral tumor detection method according to claim 1, characterized in that: The determining, based on the potential diffusion range, a pathological state corresponding to the oral cavity of the patient to be detected includes: Analyzing diffusion characteristic information corresponding to the potential diffusion range; Based on the diffusion characteristic information, query a preset medical case database to obtain diffusion-similar cases; Compare the diffusion similar cases with the current state of the patient to be detected, and extract the corresponding common features; Based on the common features, combined with preset professional pathological knowledge, a preliminary judgment is made to form a pathological hypothesis corresponding to the oral cavity of the patient to be tested; Based on the pathological hypothesis, a pathological state corresponding to the oral cavity of the patient to be tested is determined.

10. An intelligent oral tumor detection system, characterized in that: A system for executing an intelligent oral tumor detection method according to any one of claims 1 to 9, the system comprising: A signal calibration module is used to collect basic oral information of the patient to be detected, perform a layered scan on the oral cavity of the patient to be detected based on the basic oral information to obtain layered scan data, detect an oral cavity scan signal corresponding to the layered scan data, perform signal denoising on the oral cavity scan signal to obtain a denoised oral cavity signal, and perform calibration processing on the denoised oral cavity signal to obtain an oral cavity calibration signal; a change pattern analysis module, for generating an oral feature set corresponding to the patient's oral cavity based on the oral calibration signal, screening core parameters indicating tumors according to the oral feature set combined with a preset large-scale oral tumor database, and analyzing the oral change pattern corresponding to the oral cavity of the patient to be tested based on the core parameters; A system evaluation module is used to mine the degree of oral neoplasia under the oral change mode, analyze the quantitative relationship corresponding to the degree of oral neoplasia, combine the quantitative relationship with the oral state of the patient to be tested, and construct an oral neoplasia evaluation system corresponding to the patient to be tested; a diffusion range calculation module, for determining the suspected tumor area corresponding to the oral cavity of the patient to be tested based on the oral neoplasia assessment system, identifying the regional features corresponding to the suspected tumor area, analyzing the development trend of the suspected tumor area based on the regional features, and calculating the potential diffusion range corresponding to the suspected tumor area based on the development trend; A report generation module is used to determine the pathological state corresponding to the oral cavity of the patient to be detected based on the potential spread range, analyze the state feature cluster corresponding to the pathological state, judge the oral tumor risk level corresponding to the patient to be detected based on the state feature cluster, extract the key pathogenic factors in the oral tumor risk level, and generate an oral tumor detection report for the patient to be detected based on the key pathogenic factors.

Images