AI-driven personalized oral treatment scheme generation system and method

Abstract

The invention discloses an AI-driven personalized oral treatment scheme generation system and method, and particularly relates to the technical field of artificial intelligence. Multi-modal oral related data of a patient is obtained by controlling an image data acquisition unit and a non-image data input unit; the data is subjected to format unification and preprocessing through a data standardization unit, received multi-modal data is processed through a deep learning network, deep fusion of different-modal data is achieved, high-dimensional features capable of representing the oral cavity condition of a patient are extracted, and the oral cavity condition of the patient is determined based on the extracted fusion features. The oral disease diagnosis unit is used for accurately diagnosing oral diseases of a patient, generating a plurality of sets of treatment sequence schemes aiming at individual requirements of the patient, quantitatively evaluating each sequence scheme in combination with an expected utility function, receiving the output individual treatment schemes, and visually displaying the individual treatment schemes to a doctor and the patient through the visual display unit. And the doctor-patient collaborative decision-making unit is utilized to support the doctor to modify the scheme to generate a final treatment scheme.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology, and more specifically, to an AI-driven system and method for generating personalized oral treatment plans. Background Technology

[0002] With the rapid development of medical technology, oral diseases are increasingly becoming a significant issue affecting human health. From common dental caries and periodontal disease to complex malocclusions and maxillofacial tumors, each patient's oral condition exhibits a high degree of individual variability.

[0003] In current technology, dentists typically rely on their personal clinical experience, professional knowledge, and judgment based on limited examination results when developing treatment plans. For example, in the treatment of periodontal disease, dentists will use indicators such as the degree of bone resorption and periodontal pocket depth shown on X-rays, combined with clinical examination results, to determine the severity of periodontitis and formulate treatment plans such as scaling and flap surgery. Although experienced dentists can make relatively reasonable judgments, this method has obvious limitations: it is difficult to fully integrate and utilize massive amounts of multimodal data, resulting in an insufficient understanding of individual patient differences; secondly, when dealing with complex cases or rare diseases, the experience of a single dentist may not be sufficient to cover all situations. Summary of the Invention

[0004] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide an AI-driven personalized oral treatment plan generation system and method to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: an AI-driven personalized oral treatment plan generation system, comprising a data acquisition module, a feature fusion module, a disease diagnosis module, a personalized treatment plan generation module, and a treatment plan output module; The data acquisition module includes an image data acquisition unit, a non-image data input unit, and a data standardization unit. It acquires multimodal oral cavity-related data of the patient by controlling the image data acquisition unit and the non-image data input unit, and sends the data to the feature fusion module after unifying the format and preprocessing it by the data standardization unit. The feature fusion module uses a deep learning network to process the received multimodal data, achieve deep fusion of different modal data, and extract high-dimensional features that can characterize the patient's oral condition. The disease diagnosis module is based on the fusion features extracted by the feature fusion module to accurately diagnose the patient's oral diseases; The personalized treatment plan generation module generates multiple treatment sequence plans tailored to the individual needs of patients based on the diagnostic results of the disease diagnosis module and a preset treatment knowledge base, and quantitatively evaluates each sequence plan in conjunction with the expected utility function. The treatment plan output module includes a visualization display unit and a doctor-patient collaborative decision-making unit. It receives the output personalized treatment plan and displays it intuitively to doctors and patients through the visualization display unit. The doctor-patient collaborative decision-making unit supports doctors in modifying the plan and generating the final treatment plan.

[0006] Preferably, as a preferred embodiment of the AI-driven personalized oral treatment plan generation system of the present invention, the data acquisition module includes an image data acquisition unit, a non-image data input unit, and a data standardization unit. The system acquires multimodal oral-related data of the patient by controlling the image data acquisition unit and the non-image data input unit, and after the data standardization unit performs format unification and preprocessing, it sends the data to the feature fusion module. Specifically, this includes: The image data acquisition unit receives and acquires various oral imaging data of the patient, including two-dimensional images, three-dimensional images, intraoral scan images, and extraoral photographic images. The non-image data input unit receives and acquires non-image oral-related data of the patient from different information sources, including electronic medical record data, genomic data, biomarker data, and comprehensive information of the patient, including basic demographic information, overall health status, and oral-specific information. Based on the diagnostic results, the acquired data is transmitted to the data standardization unit. The data standardization unit receives raw data from the image data acquisition unit and non-image data input unit, and preprocesses the image data, including format conversion, image enhancement, registration and 3D reconstruction. It eliminates the influence of dimensions by scaling numerical data of different dimensions, encapsulates the preprocessed multimodal data, and sends it to the feature fusion module.

[0007] Preferably, as a preferred embodiment of the AI-driven personalized oral treatment plan generation system of the present invention, the feature fusion module utilizes a deep learning network to process the received multimodal data, achieving deep fusion of different modal data and extracting high-dimensional features that can characterize the patient's oral condition, specifically including: For the received image data, a convolutional neural network (CNN) is used to process it, extracting low- to high-level visual features related to the oral cavity from the raw image data; for the received non-image data, feature extraction is performed using the corresponding deep learning network according to its data type. For structured numerical data, a fully connected neural network is used; for text data, a natural language processing model is used; for genomic data, a specialized gene sequence analysis network is used to extract features related to the patient's medical history, genes, and treatment response. Features extracted from different modalities are deeply fused through an attention mechanism to form a unified high-dimensional feature representation that can comprehensively characterize the patient's oral condition.

[0008] Preferably, as a preferred embodiment of the AI-driven personalized oral treatment plan generation system of the present invention, the disease diagnosis module is used to accurately diagnose the patient's oral diseases based on the fusion features extracted by the feature fusion module, specifically including: The received high-dimensional fused feature vector is used as input to a deep learning diagnostic model. This model utilizes its learned complex patterns and associations to perform deep reasoning on the input features, outputting a diagnostic result for the patient's oral disease, as well as a probability value for each diagnostic category. Further, it includes: The high-dimensional fusion feature vector is represented as Where n is the feature dimension, a multilayer perceptron model is used. The received high-dimensional fused feature vector x is passed to the input layer of the model, processed through the hidden layer, and the output of the neuron in the output layer is obtained. The class probability is calculated using the Softmax function: ,in, Let represent the probability that the input feature x belongs to category c, and y represent the category label variable. and These are the unnormalized output values ​​corresponding to categories c and j, respectively, where C is the total number of disease categories; Based on the probability values, the model diagnoses the patient's oral diseases, and the diagnosis result is the category with the highest probability: The probability for each category Set a threshold ,when This indicates that the patient belongs to this category; otherwise, this category is not applicable. The final accurate diagnosis results, including disease category, severity, and probability value, are output to the personalized treatment plan generation module.

[0009] Preferably, as a preferred embodiment of the AI-driven personalized oral treatment plan generation system of the present invention, the personalized treatment plan generation module generates multiple treatment sequence plans tailored to the individual needs of patients based on the diagnostic results of the disease diagnosis module and a preset treatment knowledge base, and quantitatively evaluates each sequence plan in conjunction with an expected utility function, specifically including: Receive accurate diagnostic results from the disease diagnosis module, including the determined disease category, severity and probability value, and integrate comprehensive patient information, including basic demographic information, overall health status and oral-specific information. Based on the diagnostic results, perform intelligent retrieval in the preset treatment knowledge base, and combine the patient's diagnostic results and the initially integrated patient information to select a preliminary treatment plan from the knowledge base. The initially generated treatment plan is personalized and optimized, including adjustments based on overall health status, patient preferences, and oral-specific information. Using a deep reinforcement learning model, different treatment steps are treated as actions, and changes in patient status are treated as environmental feedback. Through simulation and learning, the treatment sequence is dynamically optimized to generate the best treatment plan. After personalized adjustments, multiple candidate treatment sequence plans are generated. For each candidate treatment sequence plan generated by the deep reinforcement learning model, a comprehensive expected utility function is constructed to quantitatively evaluate the treatment success rate, patient comfort, and treatment cost of each sequence plan.

[0010] Preferably, as a preferred embodiment of the AI-driven personalized oral treatment plan generation system of the present invention, the treatment plan output module includes a visualization display unit and a doctor-patient collaborative decision-making unit. The module receives the output personalized treatment plan and displays it intuitively to the doctor and patient through the visualization display unit. The doctor-patient collaborative decision-making unit supports the doctor in modifying the plan and generating the final treatment plan, specifically including: The visualization unit receives the output personalized treatment plan and displays summary information of all candidate plans in the form of side-by-side cards. Each card highlights the name / number of the plan, the estimated treatment success rate, the estimated patient comfort level, and the estimated treatment cost. The doctor-patient collaborative decision-making unit selects a candidate solution from the visualization unit as the starting point for collaborative decision-making, adjusts the execution order of treatment steps based on the doctor's clinical experience, and modifies any parameters of the solution.

[0011] This application also provides an AI-driven method for generating personalized oral treatment plans, the method specifically including: Multimodal oral cavity-related data of patients are acquired by controlling the image data acquisition unit and the non-image data input unit, and the data is formatted and preprocessed by the data standardization unit. The received multimodal data is processed using a deep learning network to achieve deep fusion of different modal data and extract high-dimensional features that can characterize the patient's oral condition. Based on the extracted fusion features, accurate diagnosis of patients' oral diseases can be made; Based on the diagnostic results and the pre-set treatment knowledge base, multiple treatment sequence plans are generated to meet the individual needs of patients, and the expected utility function is used to quantitatively evaluate each sequence plan. The system receives personalized treatment plans and presents them intuitively to doctors and patients through a visualization unit. It also supports doctors in modifying the plans using a collaborative decision-making unit to generate the final treatment plan.

[0012] On the other hand, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, wherein when the computer program is executed by the processor, it implements the functional modules of an AI-driven personalized oral treatment plan generation system as described above.

[0013] On the other hand, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements an AI-driven personalized oral treatment plan generation system as described above.

[0014] The technical effects and advantages provided by the present invention in the above technical solution are as follows: By deeply fusing imaging and non-imaging data, high-dimensional features of a patient's oral condition can be extracted, helping doctors make more accurate diagnoses and develop personalized treatment plans, thereby reducing the waste of medical resources and lowering patients' medical expenses. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0016] Figure 1 This is a flowchart of the method of the present invention.

[0017] Figure 2 This is a diagram of a multimodal feature fusion network architecture.

[0018] Table 1 is a data recording table of the simulation experiment of this invention. Detailed Implementation

[0019] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0020] In the description of this application, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0021] In the description of this application, the term "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed in this application.

[0022] Example 1 This embodiment provides, for example Figure 1 The system shown is an AI-driven personalized oral treatment plan generation system, which specifically includes a data acquisition module, a feature fusion module, a disease diagnosis module, a personalized treatment plan generation module, and a treatment plan output module. The data acquisition module includes an image data acquisition unit, a non-image data input unit, and a data standardization unit. It acquires multimodal oral cavity-related data of the patient by controlling the image data acquisition unit and the non-image data input unit, and sends the data to the feature fusion module after unifying the format and preprocessing it by the data standardization unit. The feature fusion module uses a deep learning network to process the received multimodal data, achieve deep fusion of different modal data, and extract high-dimensional features that can characterize the patient's oral condition. The disease diagnosis module is based on the fusion features extracted by the feature fusion module to accurately diagnose the patient's oral diseases; The personalized treatment plan generation module generates multiple treatment sequence plans tailored to the individual needs of patients based on the diagnostic results of the disease diagnosis module and a preset treatment knowledge base, and quantitatively evaluates each sequence plan in conjunction with the expected utility function. The treatment plan output module includes a visualization display unit and a doctor-patient collaborative decision-making unit. It receives the output personalized treatment plan and displays it intuitively to doctors and patients through the visualization display unit. The doctor-patient collaborative decision-making unit supports doctors in modifying the plan and generating the final treatment plan.

[0023] In this embodiment, the data acquisition module is specifically described. The data acquisition module includes an image data acquisition unit, a non-image data input unit, and a data standardization unit. It acquires multimodal oral cavity-related data of the patient by controlling the image data acquisition unit and the non-image data input unit. After the data is formatted and preprocessed by the data standardization unit, it is sent to the feature fusion module. Specifically, this includes: The image data acquisition unit receives and acquires various oral imaging data of the patient, including two-dimensional images, three-dimensional images, intraoral scan images, and extraoral photographic images. The non-image data input unit receives and acquires non-image oral-related data of the patient from different information sources, including electronic medical record data, genomic data, biomarker data, and comprehensive information of the patient, including basic demographic information, overall health status, and oral-specific information. Based on the diagnostic results, the acquired data is transmitted to the data standardization unit. The data standardization unit receives raw data from the image data acquisition unit and non-image data input unit, and preprocesses the image data, including format conversion, image enhancement, registration and 3D reconstruction. It eliminates the influence of dimensions by scaling numerical data of different dimensions, encapsulates the preprocessed multimodal data, and sends it to the feature fusion module.

[0024] In this embodiment, the feature fusion module is specifically described. This module utilizes a deep learning network to process the received multimodal data, achieving deep fusion of different modalities and extracting high-dimensional features that characterize the patient's oral cavity condition. Specifically, it includes: For the received image data, a convolutional neural network (CNN) is used to process it, extracting low- to high-level visual features related to the oral cavity from the raw image data; for the received non-image data, feature extraction is performed using the corresponding deep learning network according to its data type. For structured numerical data, a fully connected neural network is used; for text data, a natural language processing model is used; for genomic data, a specialized gene sequence analysis network is used to extract features related to the patient's medical history, genes, and treatment response. Features extracted from different modalities are deeply fused through an attention mechanism to form a unified high-dimensional feature representation that can comprehensively characterize the patient's oral condition.

[0025] In this embodiment, the disease diagnosis module is specifically described. This module, based on the fusion features extracted by the feature fusion module, performs accurate diagnosis of the patient's oral diseases, specifically including: The received high-dimensional fused feature vector is used as input to a deep learning diagnostic model. This model utilizes its learned complex patterns and associations to perform deep reasoning on the input features, outputting a diagnostic result for the patient's oral disease, as well as a probability value for each diagnostic category. Further, it includes: The high-dimensional fusion feature vector is represented as Where n is the feature dimension, a multilayer perceptron model is used. The received high-dimensional fused feature vector x is passed to the input layer of the model, processed through the hidden layer, and the output of the neuron in the output layer is obtained. The class probability is calculated using the Softmax function: ,in, Let represent the probability that the input feature x belongs to category c, and y represent the category label variable. and These are the unnormalized output values ​​corresponding to categories c and j, respectively, where C is the total number of disease categories; Based on the probability values, the model diagnoses the patient's oral diseases, and the diagnosis result is the category with the highest probability: The probability for each category Set a threshold ,when This indicates that the patient belongs to this category; otherwise, this category is not applicable. The final accurate diagnosis results, including disease category, severity, and probability value, are output to the personalized treatment plan generation module.

[0026] In this embodiment, the personalized treatment plan generation module is specifically described. This module generates multiple treatment sequence plans tailored to the individualized needs of patients based on the diagnostic results from the disease diagnosis module and a preset treatment knowledge base. It then quantitatively evaluates each sequence plan using an expected utility function. Specifically, this includes: Receive accurate diagnostic results from the disease diagnosis module, including the determined disease category, severity and probability value, and integrate comprehensive patient information, including basic demographic information, overall health status and oral-specific information. Based on the diagnostic results, perform intelligent retrieval in the preset treatment knowledge base, and combine the patient's diagnostic results and the initially integrated patient information to select a preliminary treatment plan from the knowledge base. The initially generated treatment plan is personalized and optimized, including adjustments based on overall health status, patient preferences, and oral-specific information. Using a deep reinforcement learning model, different treatment steps are treated as actions, and changes in patient status are treated as environmental feedback. Through simulation and learning, the treatment sequence is dynamically optimized to generate the best treatment plan. After personalized adjustments, multiple candidate treatment sequence plans are generated. For each candidate treatment sequence plan generated by the deep reinforcement learning model, a comprehensive expected utility function is constructed to quantitatively evaluate the treatment success rate, patient comfort, and treatment cost of each sequence plan.

[0027] In this embodiment, the treatment plan output module is specifically described. This module includes a visualization unit and a doctor-patient collaborative decision-making unit. It receives the output personalized treatment plan and displays it intuitively to the doctor and patient through the visualization unit. The doctor-patient collaborative decision-making unit supports doctors in modifying the plan and generating the final treatment plan, specifically including: The visualization unit receives the output personalized treatment plan and displays summary information of all candidate plans in the form of side-by-side cards. Each card highlights the name / number of the plan, the estimated treatment success rate, the estimated patient comfort level, and the estimated treatment cost. The doctor-patient collaborative decision-making unit selects a candidate solution from the visualization unit as the starting point for collaborative decision-making, adjusts the execution order of treatment steps based on the doctor's clinical experience, and modifies any parameters of the solution.

[0028] Example 2 This application also provides an AI-driven method for generating personalized oral treatment plans, the method specifically including: Multimodal oral cavity-related data of patients are acquired by controlling the image data acquisition unit and the non-image data input unit, and the data is formatted and preprocessed by the data standardization unit. The received multimodal data is processed using a deep learning network to achieve deep fusion of different modal data and extract high-dimensional features that can characterize the patient's oral condition. Based on the extracted fusion features, accurate diagnosis of patients' oral diseases can be made; Based on the diagnostic results and the pre-set treatment knowledge base, multiple treatment sequence plans are generated to meet the individual needs of patients, and the expected utility function is used to quantitatively evaluate each sequence plan. The system receives personalized treatment plans and presents them intuitively to doctors and patients through a visualization unit. It also supports doctors in modifying the plans using a collaborative decision-making unit to generate the final treatment plan.

[0029] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0030] This embodiment also provides a storage medium storing a computer program. When executed by a processor, the program implements the functional modules of an AI-driven personalized oral treatment plan generation system as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0031] Example 3 The following is another embodiment of the present invention, which provides an AI-driven personalized oral treatment plan generation system and method. In order to verify the beneficial effects of the present invention, a simulation experiment is conducted for scientific demonstration.

[0032] This experiment aims to verify the effectiveness of the AI-driven personalized oral treatment plan generation system. It primarily utilizes data collection, feature fusion, disease diagnosis, personalized treatment plan generation, and treatment plan output technologies to improve the accuracy and personalization of oral treatment plans. The experiment uses simulated and actual collected oral data, including patients' oral imaging data, non-imaging data, treatment plans, and their expected outcomes. By analyzing the degree of consistency between the generated treatment plans and the patients' actual oral conditions, the accuracy and reliability of the system in generating personalized treatment plans are verified. The simulation experiment steps are implemented according to the content of the AI-driven personalized oral treatment plan generation system provided in Example 1, and the specific steps include: The patient's oral imaging data and non-imaging data are collected, and the data are uniformly formatted and preprocessed through data standardization units. The collection frequency is set to be once before each treatment to ensure the timeliness and comprehensiveness of the data. Deep learning networks are used to process received oral imaging data and non-imaging data to achieve deep fusion of different modalities and extract high-dimensional features to more accurately represent the patient's oral condition. Based on the fusion features extracted by the feature fusion module, an AI diagnostic model is used to accurately diagnose patients' oral diseases and generate preliminary treatment plans based on the diagnostic results. Based on the diagnostic results from the diagnostic module and the preset treatment knowledge base, multiple treatment plans tailored to the individual needs of patients are generated, and the expected effects of each treatment plan are evaluated. The generated personalized treatment plan is displayed through the treatment plan output module, including the details, steps, and expected effects of the treatment plan. This allows doctors to modify the treatment plan based on patient feedback and the actual situation, and generate the final treatment plan.

[0033] The specific data from the above simulation experiment are as follows: Time / minute Image data Non-image data Diagnostic results Risk assessment results Recommended treatment plan Treatment plan modification Final treatment plan 0-5 Normal dental images No history of allergies, normal lifestyle. healthy Low risk Regular cleaning no Regular cleaning 5-10 Cavity images Excessive sugar intake and lack of regular check-ups Tooth decay Medium risk dental filling treatment no dental filling treatment 10-15 Gingivitis imaging Smoking history, improper oral care Gingivitis High risk Gingival cleaning treatment yes Gingival cleaning treatment 15-20 Periodontal disease imaging Neglecting oral health for a long time, genetic factors Periodontal disease High risk Periodontal treatment yes Periodontal treatment 20-25 Image of missing teeth Improper eating habits can lead to tooth displacement. Tooth loss High risk Dental implants yes Dental implants 25-30 Normal dental images Regular check-ups, no obvious diseases. healthy Low risk Regular inspection no Regular inspection Table 1 Experimental Analysis: By comparing the consistency between the generated treatment plans and the actual treatment results of patients, the accuracy and reliability of the system in generating personalized oral treatment plans were verified. The experimental results show that the AI-driven personalized oral treatment plan generation system can effectively improve the personalization of treatment, optimize the patient's treatment experience, and adjust the treatment plan in a timely manner according to the patient's oral health status, providing real-time and accurate treatment support.

[0034] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. An AI-driven personalized oral treatment plan generation system, characterized in that: It includes a data acquisition module, a feature fusion module, a disease diagnosis module, a personalized treatment plan generation module, and a treatment plan output module; The data acquisition module includes an image data acquisition unit, a non-image data input unit, and a data standardization unit. It acquires multimodal oral cavity-related data of the patient by controlling the image data acquisition unit and the non-image data input unit, and sends the data to the feature fusion module after unifying the format and preprocessing it by the data standardization unit. The feature fusion module uses a deep learning network to process the received multimodal data, achieve deep fusion of different modal data, and extract high-dimensional features that can characterize the patient's oral condition. The disease diagnosis module is based on the fusion features extracted by the feature fusion module to accurately diagnose the patient's oral diseases; The personalized treatment plan generation module generates multiple treatment sequence plans tailored to the individual needs of patients based on the diagnostic results of the disease diagnosis module and a preset treatment knowledge base, and quantitatively evaluates each sequence plan in conjunction with the expected utility function. The treatment plan output module includes a visualization display unit and a doctor-patient collaborative decision-making unit. It receives the output personalized treatment plan and displays it intuitively to doctors and patients through the visualization display unit. The doctor-patient collaborative decision-making unit supports doctors in modifying the plan and generating the final treatment plan.

2. The AI-driven personalized oral treatment plan generation system according to claim 1, characterized in that: The data acquisition module includes an image data acquisition unit, a non-image data input unit, and a data standardization unit, specifically including: The image data acquisition unit receives and acquires various oral imaging data of the patient, including two-dimensional images, three-dimensional images, intraoral scan images, and extraoral photographic images. The non-image data input unit receives and acquires non-image oral-related data of the patient from different information sources, including electronic medical record data, genomic data, biomarker data, and comprehensive information of the patient, including basic demographic information, overall health status, and oral-specific information. Based on the diagnostic results, the acquired data is transmitted to the data standardization unit. The data standardization unit receives raw data from the image data acquisition unit and non-image data input unit, and preprocesses the image data, including format conversion, image enhancement, registration and 3D reconstruction. It eliminates the influence of dimensions by scaling numerical data of different dimensions, encapsulates the preprocessed multimodal data, and sends it to the feature fusion module.

3. The AI-driven personalized oral treatment plan generation system according to claim 1, characterized in that: The feature fusion module utilizes a deep learning network to process the received multimodal data, achieving deep fusion of different modalities and extracting high-dimensional features that characterize the patient's oral condition. Specifically, this includes: For the received image data, a convolutional neural network (CNN) is used to process it, extracting low- to high-level visual features related to the oral cavity from the raw image data; for the received non-image data, feature extraction is performed using the corresponding deep learning network according to its data type. For structured numerical data, a fully connected neural network is used; for text data, a natural language processing model is used; for genomic data, a specialized gene sequence analysis network is used to extract features related to the patient's medical history, genes, and treatment response. Features extracted from different modalities are deeply fused through an attention mechanism to form a unified high-dimensional feature representation that can comprehensively characterize the patient's oral condition.

4. The AI-driven personalized oral treatment plan generation system according to claim 1, characterized in that: The disease diagnosis module is based on the fusion features extracted by the feature fusion module to accurately diagnose the patient's oral diseases, specifically including: The received high-dimensional fusion feature vector is used as input to a deep learning diagnostic model. The diagnostic model uses the complex patterns and associations it has learned to perform deep reasoning on the input features and outputs the diagnostic results for the patient's oral disease, as well as the probability value of each diagnostic category. The final accurate diagnosis results, including disease category, severity, and probability value, are output to the personalized treatment plan generation module.

5. The AI-driven personalized oral treatment plan generation system according to claim 4, characterized in that: The process of performing deep reasoning on the input features to output the diagnostic results for the patient's oral disease, along with a probability value for each diagnostic category, further includes: The high-dimensional fusion feature vector is represented as Where n is the feature dimension, a multilayer perceptron model is used. The received high-dimensional fused feature vector x is passed to the input layer of the model, processed through the hidden layer, and the output of the neuron in the output layer is obtained. The class probability is calculated using the Softmax function: ,in, Let represent the probability that the input feature x belongs to category c, and y represent the category label variable. and These are the unnormalized output values ​​corresponding to categories c and j, respectively, where C is the total number of disease categories; Based on the probability values, the model diagnoses the patient's oral diseases, and the diagnosis result is the category with the highest probability: The probability for each category Set a threshold ,when This indicates that the patient belongs to this category; otherwise, this category is not applicable.

6. The AI-driven personalized oral treatment plan generation system according to claim 1, characterized in that: The personalized treatment plan generation module generates multiple treatment sequence plans tailored to the individualized needs of patients based on the diagnostic results of the disease diagnosis module and a preset treatment knowledge base. It then quantitatively evaluates each sequence plan using an expected utility function, specifically including: Receive accurate diagnostic results from the disease diagnosis module, including the determined disease category, severity and probability value, and integrate comprehensive patient information, including basic demographic information, overall health status and oral-specific information. Based on the diagnostic results, perform intelligent retrieval in the preset treatment knowledge base, and combine the patient's diagnostic results and the initially integrated patient information to select a preliminary treatment plan from the knowledge base. The initially generated treatment plan is personalized and optimized, including adjustments based on overall health status, patient preferences, and oral-specific information. Using a deep reinforcement learning model, different treatment steps are treated as actions, and changes in patient status are treated as environmental feedback. Through simulation and learning, the treatment sequence is dynamically optimized to generate the best treatment plan. After personalized adjustments, multiple candidate treatment sequence plans are generated. For each candidate treatment sequence plan generated by the deep reinforcement learning model, a comprehensive expected utility function is constructed to quantitatively evaluate the treatment success rate, patient comfort, and treatment cost of each sequence plan.

7. The AI-driven personalized oral treatment plan generation system according to claim 1, characterized in that: The treatment plan output module includes a visualization unit and a doctor-patient collaborative decision-making unit. It receives the output personalized treatment plan and displays it intuitively to doctors and patients through the visualization unit. The doctor-patient collaborative decision-making unit supports doctors in modifying the plan and generating the final treatment plan, specifically including: The visualization unit receives the output personalized treatment plan and displays summary information of all candidate plans in the form of side-by-side cards. Each card highlights the name / number of the plan, the estimated treatment success rate, the estimated patient comfort level, and the estimated treatment cost. The doctor-patient collaborative decision-making unit selects a candidate solution from the visualization unit as the starting point for collaborative decision-making, adjusts the execution order of treatment steps based on the doctor's clinical experience, and modifies any parameters of the solution.

8. An AI-driven personalized oral treatment plan generation method applied to an AI-driven personalized oral treatment plan generation system as described in any one of claims 1-7, characterized in that: Specifically, it includes: Multimodal oral cavity-related data of patients are acquired by controlling the image data acquisition unit and the non-image data input unit, and the data is formatted and preprocessed by the data standardization unit. The received multimodal data is processed using a deep learning network to achieve deep fusion of different modal data and extract high-dimensional features that can characterize the patient's oral condition. Based on the extracted fusion features, accurate diagnosis of patients' oral diseases can be made; Based on the diagnostic results and the pre-set treatment knowledge base, multiple treatment sequence plans are generated to meet the individual needs of patients, and the expected utility function is used to quantitatively evaluate each sequence plan. The system receives personalized treatment plans and presents them intuitively to doctors and patients through a visualization unit. It also supports doctors in modifying the plans using a collaborative decision-making unit to generate the final treatment plan.

Images