A treatment decision support system for patients with advanced non-small cell lung cancer

Abstract

The application belongs to the auxiliary technical field of treatment decision, and specifically discloses a treatment decision auxiliary system for patients with advanced non-small cell lung cancer, which comprises a platform server, a medical terminal, a patient terminal, a data server and a communication network; based on the pre-test scores of the decision preparation scale and the decision dilemma scale uploaded by the patient terminal, the dimension with an abnormal score is marked, and a root cause dimension is determined; the education module is extracted in a manner that the education module corresponding to the root cause dimension occupies a dominant number; based on the learning behavior data and the knowledge test accuracy, the knowledge acquisition degree index is determined when it is determined that effective learning is completed; then, the confidence level of the patient's preference score for each treatment scheme is determined; and a structured treatment decision auxiliary report is generated. Through the application, the root cause of the decision dilemma of the patient can be identified under the premise that the patient is fully informed, personalized education content is pushed on demand, and the preference reliability is quantitatively evaluated, so as to assist the patient in participating in the treatment decision.

Description

Technical Field

[0001] This application belongs to the field of treatment decision support technology, and more specifically, relates to a treatment decision support system for patients with advanced non-small cell lung cancer. Background Technology

[0002] Non-small cell lung cancer (NSCLC) is the most common pathological type of lung cancer, accounting for approximately 80% of all lung cancers. Due to its insidious symptoms, about 70% of NSCLC patients are already in the advanced stages when they seek medical attention. With advancements in modern medical technology, treatment decisions for patients with advanced NSCLC have become more complex and diverse, encompassing various treatment options such as chemotherapy, radiotherapy, targeted therapy, and immunotherapy. However, factors such as a lack of medical information, unclear patient values, limited doctor-patient communication, and insufficient social support can make it difficult for patients and their families to determine which treatment option to choose, leading to a decision-making dilemma.

[0003] Related studies show that approximately 89% of lung cancer patients experience decision-making difficulties to varying degrees. These difficulties can lead to negative emotions such as anxiety and depression, resulting in unsatisfactory decisions, delayed treatment, decreased quality of life, and strained doctor-patient relationships. Therefore, alleviating decision-making difficulties for patients with advanced non-small cell lung cancer and their families is an important issue that needs to be addressed in clinical practice.

[0004] Decision support, as a tool for shared decision-making, helps patients and their families understand the risks and benefits of each treatment option by providing them with information related to their disease, and helps patients clarify their own values ​​and preferences. Numerous studies both domestically and internationally have confirmed that the application of decision support tools in clinical practice can effectively reduce decision-making difficulties for cancer patients, improve decision satisfaction, and enhance doctor-patient communication.

[0005] Treatment decisions for advanced non-small cell lung cancer are characterized by diverse options, complex value trade-offs, and significant differences in patients' level of informed consent. For this treatment decision-making scenario, identifying the root causes of patients' decision-making dilemmas while ensuring they are fully informed, delivering personalized educational content as needed, and quantitatively assessing the credibility of their preferences to assist patients in participating in treatment decisions are urgent technical challenges that need to be addressed in this field. Summary of the Invention

[0006] To address the shortcomings of existing technologies, the purpose of this application is to assist patients in participating in treatment decisions by identifying the root causes of their decision-making difficulties with full informed consent, delivering personalized educational content as needed, and quantitatively assessing the credibility of their preferences, thereby addressing the treatment decision-making scenarios of patients with advanced non-small cell lung cancer.

[0007] To achieve the above objectives, in a first aspect, this application provides a treatment decision support system for patients with advanced non-small cell lung cancer, comprising: a platform server, a medical terminal, a patient terminal, a data server, and a communication network, wherein the platform server is connected to the medical terminal, the patient terminal, and the data server via the communication network. The patient terminal is used to collect basic patient information and relevant data in the decision-making process, display treatment plan information and educational content, guide patients to complete questionnaires, knowledge learning, knowledge tests and treatment plan preference ratings, and upload the collected data to the platform server. The platform server is used to standardize the pre-test scores of the decision readiness scale and decision dilemma scale uploaded by patients, compare them with preset thresholds, and mark the dimensions with abnormal scores. Based on the marked abnormal dimensions, a root cause dimension is determined. Based on the abnormal dimensions, a dimension-content mapping table is queried, and educational modules are extracted according to the number of educational modules corresponding to the root cause dimension that dominate the overall learning path. These modules are then combined into personalized learning paths and pushed to the patient's terminal. Based on the patient's learning behavior data and knowledge test accuracy in the educational modules, it is determined whether the patient has completed effective learning. If effective learning is determined, the patient's knowledge acquisition level index for each treatment plan is determined. Based on the knowledge acquisition level index, the confidence level and confidence score of the patient's preference rating for each treatment plan are determined. Based on the knowledge acquisition level index of the treatment plan finally selected by the patient, the patient's preferences are stratified and labeled, and a structured treatment decision support report is generated. The medical terminal is used to receive treatment decision support reports and hierarchical labels pushed by the platform server, so that medical staff can view patient information and enter the final treatment plan. The data server is used to store patient data sent by the platform server after it has been anonymized, and to classify, archive, and manage the anonymized data.

[0008] In one possible implementation, the platform server labels dimensions of score anomalies, including: For each dimension in the decision readiness measure, based on the standardized score of the dimension, if the standardized score is lower than the preset outlier threshold of the decision readiness measure, the dimension is marked as an outlier dimension; for each dimension in the decision dilemma measure, based on the standardized score of the dimension, if the standardized score is higher than the preset outlier threshold of the decision dilemma measure, the dimension is marked as an outlier dimension. Based on the labeled outlier dimensions, the deviation of each outlier dimension is calculated; for the decision readiness metric, the dimension deviation is obtained using the following formula: ; in, This indicates the degree of deviation from the outlier dimensions in the decision readiness measure table; This represents the preset anomaly threshold for the corresponding dimension in the decision readiness metrics table; This represents the standardized actual score of the corresponding dimension in the decision readiness metric. For the decision-making dilemma scale, the dimensional deviation is obtained using the following formula: ; in, Indicates the degree of deviation from the abnormal dimension of the decision-making dilemma scale; This represents the standardized actual score of the corresponding dimension in the Decision Dilemma Scale; This represents the preset abnormal threshold for the corresponding dimension in the Decision Dilemma Scale.

[0009] In one possible implementation, the platform server determines a root cause dimension from the anomaly dimensions, including: Based on the deviation of each anomaly dimension, the anomaly dimension with the largest absolute value of deviation is selected as the root cause dimension.

[0010] In one possible implementation, the platform server determines whether the patient has completed effective learning, including: Based on the patient's browsing time, page scrolling depth, module completion status, and knowledge quiz accuracy in each educational module, the patient's effective learning is determined after completing all pushed educational modules and knowledge quizzes. To be considered as having completed effective learning, the following conditions must be met simultaneously: ; in, This indicates the completion coverage rate of the pushed education modules. A completion coverage rate of 100% means that the patient has completed all pushed education modules. Indicates the first The scroll depth of a page with a completed education module; This indicates the patient's total study time; Indicates the first Estimated standard learning time for each educational module; This represents the sum of the estimated standard learning time for all pushed educational modules.

[0011] In one possible implementation, the platform server determines an index of the patient's knowledge acquisition level regarding various treatment options, including: Based on the tags attached to each educational module, the association between the educational module and one or more treatment plans is determined; based on the patient's correct answer rate on knowledge tests for each educational module, and combined with the association between the educational module and each treatment plan, the data is aggregated by treatment plan to obtain the patient's knowledge acquisition index for each treatment plan; where, for the first... For each treatment plan, the patient's knowledge acquisition index of the treatment plan is obtained using the following formula: ; in, Indicates the patient's... The knowledge acquisition index for each treatment plan ranges from 0 to 100. It is a positive integer; Indicates the relationship with the first A collection of educational modules associated with each treatment plan; Represents a set The serial number of the Chinese education module; This represents the patient's accuracy rate on the knowledge test in the i-th educational module, with a value ranging from 0 to 1; Indicates the relationship with the first The number of educational modules associated with each treatment plan.

[0012] In one possible implementation, the platform server determines the confidence level and confidence score of the patient's preference rating for each treatment option, including: Based on the patient's knowledge acquisition index of each treatment option, the confidence score of the patient's preference rating for each treatment option is obtained; for the first... For each treatment option, the confidence score is obtained using the following formula: ; in, Indicates the patient's... Confidence score of each treatment option preference rating; Indicates the patient's... Knowledge acquisition index of each treatment plan; Based on the A knowledge acquisition index for each treatment option is used to categorize the confidence levels of patients' treatment option preference ratings: when When is a high confidence level; when When is the middle confidence level; when The confidence level is low.

[0013] In one possible implementation, the platform servers are combined into personalized learning paths, including: Based on the anomaly dimension, query the dimension-content mapping table to obtain one or more content tags corresponding to each anomaly dimension; based on the content tags, extract the corresponding education modules from the knowledge base, and combine them into a personalized learning path in such a way that the number of education modules corresponding to the root cause dimension accounts for no less than 70% of the overall learning path, push it to the patient's terminal and send a learning reminder notification. The order of multiple educational modules under the same dimension in the personalized learning path is determined by the push priority defined in the dimension-content mapping table.

[0014] In one possible implementation, the platform server is also used for: When a patient fails to complete effective learning, remind the patient to complete the learning, and only push the educational modules that did not meet the criteria for effective learning to the patient again; Based on the patient's learning behavior data and knowledge test accuracy after completing supplementary learning, the patient's effective learning was reassessed, and the patient's knowledge acquisition index for each treatment plan was re-determined after the assessment was passed.

[0015] For example, the criteria for determining effective learning may specifically be: ,in Indicates the first The actual learning time for each educational module.

[0016] In one possible implementation, the change in scores between the pre-test and post-test is calculated; Based on the patient's condition, treatment plan preference score, scale score, changes before and after the test, confidence level, confidence score, preference stratification marker, and medical opinions, a structured treatment decision support report is generated and pushed to the medical terminal linked to the patient, where a task to be processed is generated. Based on the final treatment plan entered into the medical terminal, a final treatment decision report is generated. After anonymizing the final treatment decision report and patient privacy data, the data is sent to the data server for classified archiving and storage.

[0017] Overall, the technical solutions conceived in this application have the following beneficial effects compared with the prior art: (1) The treatment decision support system for patients with advanced non-small cell lung cancer provided in this application presents different treatment options to patients based on mobile Internet technology. The system guides patients to clarify their treatment preferences and generate decision support reports based on the actual treatment options available to each patient. Medical staff can conduct targeted communication with patients based on the content of the decision support reports, which helps to improve the efficiency of diagnosis and treatment and the quality of doctor-patient communication.

[0018] (2) This application identifies the specific difficulties of patients in the dimensions of information awareness, value clarity, social support, uncertainty and decision effectiveness through the pre- and post-test assessment mechanism of Prep DM and DCS. It also accurately determines the root cause dimension through root cause identification and pushes targeted educational content to patients in a personalized manner, which helps to improve the matching degree between decision support content and patients' actual difficulties.

[0019] (3) This application comprehensively records the patient's browsing time, scrolling depth, module completion status and knowledge test accuracy through the knowledge quantification assessment module. It uses an objective threshold to determine whether effective learning has been completed and calculates the knowledge acquisition degree index accordingly. When the patient has not completed effective learning, the application reminds the patient to supplement the learning, which helps to ensure that the patient expresses treatment preferences with full knowledge.

[0020] (4) This application combines the knowledge acquisition degree index with the confidence level through the scoring confidence calculation module and the stratification judgment module, and marks the patient preferences in three levels of high, medium and low and sends differentiated reminders to the medical staff terminal, which helps medical staff to quickly identify the credibility of the patient's preferences and take differentiated communication strategies. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the structure of the treatment decision support system for patients with advanced non-small cell lung cancer provided in the embodiments of this application; Figure 2 This is a schematic diagram of the structure of the platform server provided in the embodiments of this application; Figure 3 This is a schematic diagram of the structure of the medical terminal provided in the embodiments of this application; Figure 4 This is a schematic diagram of the structure of the patient terminal provided in the embodiments of this application; Figure 5 This is a schematic diagram of the structure of the data server provided in the embodiments of this application. Detailed Implementation

[0022] To facilitate a clearer understanding of the various embodiments of this application, some relevant background knowledge will be introduced as follows.

[0023] The realization of shared decision-making typically relies on standardized scales to quantify patients' decision-making status. The Preparation Decision Making (PrepDM) scale quantifies the degree of preparedness patients have before making treatment decisions, reflecting aspects such as the patient's awareness of disease information, clarity of their own values, and the level of social support they receive. The Decisional Conflict Scale (DCS) quantifies the intensity of the difficulties experienced by patients during the decision-making process, reflecting the level of difficulty in dimensions such as information awareness, clarity of values, social support, uncertainty, and decision effectiveness.

[0024] The meanings of the five dimensions of the above scale—information awareness, value clarity, uncertainty, support, and decision effectiveness—are explained as follows: Information awareness refers to the patient's understanding and mastery of information about the disease and the benefits, risks, and side effects of various treatment options; value clarity refers to the patient's clarity regarding their most important goals and orientations during treatment (such as prolonging survival time, maintaining quality of life, and reducing treatment side effects); uncertainty refers to the patient's degree of hesitation and lack of certainty about whether their treatment decisions are correct and appropriate; support refers to the degree of support the patient feels from family, medical staff, and society during the treatment decision-making process; and decision effectiveness refers to the degree to which the patient's treatment decisions are consistent with their own values ​​and that they are able to implement and adhere to them.

[0025] Prep DM and DCS are both interconnected and distinct: both revolve around the patient's decision-making state, and their sub-dimensions correspond and corroborate each other in areas such as informed consent, clear values, and social support; the difference lies in that Prep DM focuses on assessing the patient's readiness and preparedness before making a decision, with higher scores indicating more adequate preparation, while DCS focuses on assessing the degree of distress and conflict experienced by the patient during the decision-making process, with higher scores indicating more severe distress. Therefore, low-scoring dimensions in Prep DM and high-scoring dimensions in DCS usually point to the same type of weakness in decision-making.

[0026] Existing decision support methods often employ standardized information presentation and scoring methods, presenting the same educational content and decision-making process to all patients. This makes it difficult to personalize content recommendations based on the specific difficulties faced by each patient. Furthermore, current methods typically collect patient preference scores directly after they complete questionnaires, without quantifying the level of informed consent at the time of completion. This makes it difficult to determine whether the preferences expressed by patients are based on adequate knowledge. This background information suggests that combining questionnaire assessment, root cause identification, personalized recommendations, and informed consent quantification is a feasible direction for improving decision support outcomes for patients with advanced non-small cell lung cancer.

[0027] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0028] In the embodiments of this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design that is described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design. Specifically, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0029] In the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more, for example, multiple processing units means two or more processing units, multiple elements means two or more elements, etc.

[0030] The embodiments of this application are described below with reference to the accompanying drawings.

[0031] 1. System overall structure.

[0032] like Figure 1 As shown, this application provides a treatment decision support system 100 for patients with advanced non-small cell lung cancer, including a platform server 1, multiple medical care terminals 2, multiple patient terminals 3, a data server 4, and a communication network 5; the platform server 1 is connected to the medical care terminals 2, patient terminals 3, and data server 4 through the communication network 5.

[0033] 2. Platform server.

[0034] like Figure 2 As shown, the platform server 1 includes a patient information storage module 11, a knowledge base storage module 12, a mapping database module 13, a threshold configuration module 14, a decision evaluation module 15, a root cause identification module 16, a knowledge quantification evaluation module 17, a score confidence calculation module 18, a personalized push module 19, a decision report generation module 20, a medical and nursing task storage and distribution module 21, a hierarchical judgment module 22, a doctor-patient interaction management module 23, a patient communication review module 24, a data anonymization module 25, a platform-side communication module 26, and a platform-side control module 27.

[0035] The platform-side communication module 26 is used for data exchange between the components of the platform server 1, as well as between the platform server 1 and the medical terminal 2, the patient terminal 3, and the data server 4; the platform-side control module 27 is used to control the operation of the components of the platform server 1.

[0036] The patient information storage module 11 is used to store patient basic information, disease data, scale scores, knowledge acquisition index, treatment plan preference scores and confidence levels, knowledge test answers, doctor-patient communication records and other data.

[0037] The knowledge base storage module 12 is used to store educational content related to disease knowledge and treatment plans. The educational content includes text, interactive exercises, videos, etc. Each educational module is attached with a tag, which corresponds to the scale dimension or treatment plan type. The knowledge base storage module 12 also records the estimated standard learning time for each educational module.

[0038] The mapping database module 13 is used to store the dimension-content mapping table. The dimension-content mapping table defines the correspondence between each dimension of the scale and the knowledge content tags, as well as the push priority. The push priority is used to determine the order of multiple educational modules under the same dimension when making personalized pushes. It belongs to a different scope than the root cause identification priority used by the root cause identification module 16.

[0039] The threshold configuration module 14 is used to store preset thresholds, confidence thresholds, and effective learning judgment thresholds for each scale dimension, and supports medical staff to make custom adjustments. In this embodiment, the Prep DM abnormal threshold is 60 points, the DCS abnormal threshold is 25 points, and the rolling depth threshold and learning time ratio threshold in the effective learning judgment threshold are 80% and 60%, respectively. The above thresholds are all parameters that can be customized and adjusted by medical staff through the threshold configuration module 14, rather than fixed and immutable parameters.

[0040] The decision assessment module 15 is used to obtain the total score and scores of each dimension of the Prep DM and DCS filled in by the patient during the initial login. After standardizing the scores of each dimension, it compares them with the preset thresholds in the threshold configuration module 14 and marks the dimensions with abnormal scores. After the patient completes the decision, the decision assessment module 15 obtains the total score and scores of each dimension of the Prep DM and DCS again, and calculates the change in scores between the pre-test and post-test. The assessment at the initial login is the pre-test, and the assessment after completing the decision is the post-test. The pre-test and post-test constitute a pre-test and post-test assessment mechanism. The change is used to evaluate the intervention effect of the system on the patient's decision state, and is included in the decision support report by the decision report generation module 20 and stored by the patient information storage module 11 for subsequent effect evaluation and statistical analysis.

[0041] For example, in Prep DM, when the standardized score of a certain dimension is lower than the Prep DM outlier threshold (60 points), the dimension is marked as an outlier. For example, in DCS, when the standardized score of a certain dimension is higher than the DCS outlier threshold (25 points), the dimension is marked as an outlier.

[0042] The root cause identification module 16 is used to determine a root cause dimension from the anomalous dimensions marked by the decision evaluation module 15. The root cause identification module 16 calculates the deviation of each anomalous dimension. For Prep DM, the dimension deviation is calculated as follows: ; Parameter explanation: : Indicates the deviation of a certain abnormal dimension in Prep DM; : Indicates the preset anomaly threshold for this dimension in Prep DM; : Indicates the standardized actual score of this dimension in Prep DM.

[0043] For DCS, the dimensional deviation is calculated as follows: ; Parameter explanation: : Indicates the deviation of a certain abnormal dimension of DCS; : Represents the standardized actual score of this dimension of DCS; : Indicates the preset anomaly threshold for this dimension of DCS.

[0044] The root cause identification module 16 selects the outlier dimension with the largest absolute deviation as the root cause dimension. When two or more outlier dimensions have the same absolute deviation or the difference is within a preset tolerance range, the root cause identification module 16 determines the root cause dimension according to the root cause identification priority, which is: informed information > clear values ​​> uncertainty > support > decision effectiveness. That is, the largest absolute deviation is used to determine the root cause dimension under normal circumstances, and the root cause identification priority is only used as an auxiliary judgment criterion when the absolute deviation values ​​of multiple outlier dimensions are the same or close. If there are no outlier dimensions, the root cause dimension is marked as none.

[0045] The knowledge quantification assessment module 17 records the patient's behavioral data during the knowledge learning process. This data includes the browsing time of each educational module, page scrolling depth, module completion status, and the accuracy rate of the knowledge quiz. After the patient completes all the pushed educational modules and knowledge quizzes, the knowledge quantification assessment module 17 determines whether the patient has completed effective learning. To determine whether effective learning has been completed, the following conditions must be met simultaneously: ; Parameter explanation: This indicates the completion coverage rate of the pushed education modules. A completion coverage rate of 100% means that the patient has completed all pushed education modules. : Indicates the scroll depth of the j-th completed education module page; : Indicates the patient's total study time; : Represents the estimated standard learning time for the j-th educational module; : This represents the sum of the estimated standard learning time for all pushed educational modules.

[0046] Upon determining that effective learning has been completed, the knowledge quantification assessment module 17 determines the patient's knowledge acquisition level index for each treatment plan. In this embodiment, each educational module in the knowledge base storage module 12 is accompanied by a tag, which defines the association between the educational module and one or more treatment plans. The knowledge quantification assessment module 17 aggregates the patient's knowledge acquisition level index for each treatment plan based on the patient's correct test score on each educational module and the association between each educational module and each treatment plan. For the first... For each treatment plan, the patient's knowledge acquisition index regarding that treatment plan is calculated as follows: ; Parameter explanation: : Represents the patient's knowledge acquisition level of the m-th treatment plan, with a value ranging from 0 to 100; : Represents the set of educational modules associated with the m-th treatment plan; : Represents a set The serial number of the Chinese education module; : Represents the patient's accuracy rate on the knowledge test in the i-th educational module, with a value ranging from 0 to 1; : Indicates the number of educational modules associated with the m-th treatment plan.

[0047] When a patient fails to complete effective learning, the knowledge quantification assessment module 17 reminds the patient to supplement the learning via the notification receiving module. During supplementary learning, only the educational modules that did not meet the above-mentioned judgment criteria are re-pushed to the patient. After the patient completes the supplementary learning, the knowledge quantification assessment module 17 re-determines whether effective learning has been completed and recalculates the patient's knowledge acquisition level index for each treatment plan if the determination is passed. If the patient still fails to complete effective learning after a preset number of supplementary learning sessions, the system marks the patient's knowledge acquisition level as needing improvement and pushes this mark along with the decision support report to the medical staff terminal 2 for intervention by medical staff. The decision-making process continues after confirmation by the medical staff.

[0048] The confidence calculation module 18 is used to determine the confidence level and confidence score of the patient's preference rating for each treatment option based on the patient's knowledge acquisition index for each treatment option. For the m-th treatment option, the confidence score of the patient's preference rating for that treatment option is calculated as follows: ; Parameter explanation: : Represents the confidence score of the patient's preference rating for the m-th treatment option; : Represents the patient's level of knowledge acquisition regarding the m-th treatment option.

[0049] The confidence level of a patient's preference rating for the m-th treatment option is based on the knowledge acquisition index of that treatment option. The division is as follows: when When is a high confidence level; when When is the middle confidence level; when The confidence level is low. Confidence score. The confidence level is used to characterize the credibility of the patient's preference rating for the treatment plan with continuous numerical values. The confidence level is used to discretize the confidence score into three levels: high, medium, and low, so that medical staff can quickly identify it. The confidence calculation module 18 binds and stores the confidence level and confidence score corresponding to each treatment plan with the preference rating of the treatment plan and pushes it to the decision report generation module 20.

[0050] The personalized push module 19 is used to query the mapping database module 13 based on the root cause dimension and other abnormal dimensions determined by the root cause identification module 16, obtain one or more content tags corresponding to each dimension, and then extract the corresponding educational modules from the knowledge base storage module 12 according to the dominant number of educational modules corresponding to the root cause dimension in the overall learning path, combine them into a personalized learning path, push it to the knowledge learning module of the patient terminal 3, and send a learning reminder notification. In this embodiment, the proportion of educational modules corresponding to the root cause dimension in the overall learning path is not less than 70%, that is, the above-mentioned dominant number refers to the proportion of the number of educational modules corresponding to the root cause dimension in the overall learning path is not less than 70%, and the quantitative caliber of dominant number is consistent with this proportion.

[0051] The content ordering in personalized push notifications is determined based on the push priority defined by the mapping database module 13, while the determination of the root cause dimension is based on the principle of the maximum absolute value of the deviation of the root cause identification module 16. The two have different scopes: the root cause identification priority is used to determine which dimension is the root cause, while the push priority of the mapping database is used to determine the ordering of multiple education modules under the same dimension. Therefore, the root cause dimension does not need to be consistent with the dimension with higher priority in the mapping database, and there is no logical conflict.

[0052] The decision report generation module 20 integrates the patient's condition, treatment plan preference score, scale score, and changes in scale before and after measurement from the patient information storage module 11, as well as the confidence level, confidence score, and medical opinions output by the rating confidence calculation module 18, to generate a structured treatment decision support report.

[0053] The medical and nursing task storage and distribution module 21 is used to push the structured decision support report output by the decision report generation module 20 to the medical and nursing terminal 2 bound to the patient, and generate a task to be processed on the medical and nursing terminal 2.

[0054] The stratification judgment module 22 is used to determine the patient's level of knowledge acquisition regarding their final chosen treatment plan based on the patient's knowledge acquisition index. Patients are stratified and corresponding reminders are sent to healthcare terminal 2. The stratification rules are as follows (the threshold selected here is the same as the threshold used for the confidence level of the treatment plan preference score mentioned above): when When marked as a high-confidence preference, a routine priority reminder is sent to healthcare terminal 2, stating that the patient's preference is highly credible and can be communicated according to the routine procedure; when When marked as having a medium confidence preference, a medium-priority reminder is sent to medical staff terminal 2, suggesting that medical staff supplement communication with the patient regarding the weak points; when When a patient's preference is flagged as low-confidence and a high-priority alert is sent to healthcare terminal 2, the alert stating that the patient's preference has low credibility and requires close intervention. The hierarchical flagging and alert priority are pushed along with the decision support report, allowing healthcare staff to adopt differentiated communication strategies accordingly.

[0055] Since patients may give different ratings to different treatment options, the confidence level of patient preference is determined as follows: the confidence level of the rating corresponding to the treatment option finally chosen by the patient is used as the confidence level of the patient preference; when the patient's ratings of two or more treatment options are similar and it is difficult to distinguish the primary option, the lowest confidence score among the similar options is taken as the confidence level of the patient preference, so as to avoid overestimating the credibility of the preference.

[0056] The doctor-patient interaction management module 23 is used to establish a dedicated communication channel between the medical staff terminal 2 and the patient terminal 3, forward the patient's consultation request to the corresponding medical staff terminal 2, and push the medical staff's reply to the patient terminal 3, while recording the timestamp and content of all messages.

[0057] The patient communication review module 24 is used to receive the treatment experiences, decision-making insights, and other content posted by patients on the patient communication platform, and push the content to the task processing module of the medical terminal 2 for medical staff to verify compliance and authenticity. The patient communication review module 24 receives the approval or rejection instructions from medical staff. Approved content is publicly displayed in the patient communication area, and rejected content is notified to the patient to modify or delete it.

[0058] The data desensitization module 25 is used to desensitize the patient data stored in the platform server 1 and send the desensitized data to the data server 4 through the platform-side communication module 26.

[0059] 3. Medical and nursing terminals.

[0060] like Figure 3 As shown, the medical terminal 2 includes a patient information retrieval module 31, a doctor-patient communication module 32, a task processing module 33, a knowledge base management module 34, a statistical analysis module 35, a medical-nursing side communication module 36, and a medical-nursing side control module 37.

[0061] The medical staff communication module 36 is used for data exchange between the components of the medical staff terminal 2, as well as between the medical staff terminal 2 and the platform server 1 and the patient terminal 3; the medical staff control module 37 is used to control the operation of the components of the medical staff terminal 2.

[0062] The patient information access module 31 is used to support medical staff in viewing patients' basic information, medical records, historical scores of scales, knowledge acquisition index, score confidence level and structured decision support reports.

[0063] The doctor-patient communication module 32 is used to receive consultation requests initiated by patients, support medical staff in editing reply content, and record the complete communication history.

[0064] The task processing module 33 is used to display the tasks to be processed in a list format. The tasks to be processed include patient consultations that need to be answered, patient communication content that needs to be reviewed, and decision-making tasks that need to be confirmed.

[0065] The knowledge base management module 34 is used to support medical staff in adding, modifying, and deleting treatment plan knowledge, educational module content, scale thresholds, and dimension-content mapping relationships in the system.

[0066] The statistical analysis module 35 is used to collect statistics on the decision-making participation of a batch of patients and the system usage. The decision-making participation includes the scale completion rate, knowledge test pass rate, confidence distribution, etc., and the system usage includes the access frequency of each module, average learning time, etc.

[0067] 4. Patient terminal.

[0068] like Figure 4 As shown, the patient terminal 3 includes a user registration and login module 41, a data collection module 42, a knowledge learning module 43, a decision support module 44, a doctor-patient communication module 45, a patient communication module 46, an incentive display module 47, a decision report viewing module 48, a notification receiving module 49, a patient-side communication module 410, and a patient-side control module 411.

[0069] The patient-side communication module 410 is used for data exchange between the components of the patient terminal 3, as well as between the patient terminal 3 and the platform server 1 and the medical terminal 2; the patient-side control module 411 is used to control the operation of the components of the patient terminal 3 to ensure that the functions of each module are executed in a coordinated manner.

[0070] The user registration and login module 41 is used to support patients or their families in registering personal accounts, completing identity verification, and managing account information.

[0071] The data collection module 42 is used to collect basic information about patients and relevant data during the decision-making process. The basic information includes age, gender, pathological stage, disease type, etc. The relevant data during the decision-making process includes scale completion results, plan preference scores, knowledge test answers, etc., and the above data are synchronously uploaded to the platform server 1.

[0072] The knowledge learning module 43 is used to display educational content such as basic disease knowledge, treatment plan comparison, side effect management, and rehabilitation nursing in the form of pictures, texts, videos, etc., and supports learning progress recording, learning time statistics, and page scrolling depth tracking.

[0073] The decision support module 44 is used to guide patients to complete the standardized decision-making process preset by the platform server 1, including viewing treatment options, comparing options, scoring options preference, taking knowledge tests, and filling in pre- and post-measurement forms. The options preference score is used by patients to express their preference for each treatment option after viewing the risk-benefit comparison chart of each treatment option.

[0074] The doctor-patient communication module 45 is used to support patients in selecting specific medical staff, initiating consultation requests such as questions about treatment plans and decision-making difficulties, and receiving reply messages from medical staff.

[0075] The patient communication module 46 is used to support patients in editing and previewing their treatment experiences, decision-making insights, and other content, which is then submitted to the platform server 1 for medical staff to review and view the published content that has been approved.

[0076] The incentive display module 47 is used to display the patient's knowledge acquisition level index and learning path completion progress in real time in the form of a dashboard.

[0077] The decision report viewing module 48 is used to support patients in viewing the personal decision support reports generated by the system.

[0078] The notification receiving module 49 is used to receive notifications such as learning tasks, medical staff replies, questionnaire completion reminders, and supplementary learning tips pushed by the platform server 1.

[0079] 5. Data server.

[0080] like Figure 5As shown, the data server 4 includes a data-side communication module 51, a data-side control module 52, a de-identified data storage module 53, and a classified data management module 54.

[0081] The data-side communication module 51 is used for data exchange between the components of the data server 4, as well as between the data server 4 and the platform server 1, the medical terminal 2, and the patient terminal 3; the data-side control module 52 is used to control the operation of the components of the data server 4 to ensure the coordinated execution of data storage, desensitization, and classification.

[0082] The de-identified data storage module 53 is used to store various types of de-identified data sent from the platform server 1.

[0083] The classification data management module 54 is used to classify, archive, and retrieve the data in the desensitized data storage module 53. It divides the storage directory into treatment plan categories, patient data categories, communication content categories, and knowledge resource categories, and supports the platform server 1 to retrieve the data as needed.

[0084] 6. Examples of scale assessment and abnormal dimension determination.

[0085] In this embodiment, the decision assessment module 15 uses a decision status scale, which is used as a whole to quantify the patient's state of participation in treatment decisions. When in use, the decision status scale is further refined into two sub-scales: Prep DM and DCS, which quantify the patient's decision status from the perspectives of preparedness and predicament, respectively. An example of the decision status scale is as follows: it includes five dimensions: informed consent, clear values, social support, uncertainty, and decision effectiveness. Each dimension consists of several items. The patient scores each item using a Likert scale, and the scores are then aggregated and standardized to a range of 0 to 100.

[0086] An example of the Prep DM is as follows: The Prep DM is used to quantify a patient's overall preparedness before making a treatment decision. A higher score indicates that the patient perceives themselves to be more prepared for the decision. This scale has a single-dimensional structure, and the items cover aspects such as the patient's perception of the necessity of the decision, consideration of the advantages and disadvantages of the options, clarity of their own values, perception of their level of participation, and preparation for communication with the doctor.

[0087] An example of DCS (Distributed Risk System) is as follows: DCS is used to quantify the information dilemma, value dilemma, support dilemma, uncertainty dilemma, and decision-making ineffectiveness experienced by patients during the decision-making process. The higher the score of each dimension, the more severe the dilemma. For example, an example item of the uncertainty dimension in DCS is the patient's confidence in whether the chosen option is correct, and an example item of the support dimension is the level of support the patient feels from family and medical staff.

[0088] The connection and difference between Prep DM and DCS lies in the fact that both revolve around the patient's decision-making state and their sub-dimensions correspond to each other. Prep DM quantifies from a positive perspective of preparedness, with high scores being better, while DCS quantifies from a negative perspective of predicament, with high scores being worse. Therefore, a low score in a certain dimension of Prep DM and a high score in the corresponding dimension of DCS usually point to the same type of weak link in decision-making, and the results of the two can corroborate each other.

[0089] An example of an abnormal dimension determination is as follows: For Prep DM, assuming a patient's standardized score for the value clarity dimension is 45, which is lower than the Prep DM abnormal threshold of 60, then the value clarity dimension is marked as an abnormal dimension, with a deviation of 60. 45 = 15; For DCS, assuming a patient's standardized score for the uncertainty dimension is 40, which is higher than the DCS abnormal threshold of 25, then the uncertainty dimension is marked as an abnormal dimension with a deviation of 40. 25=15. When comparing the absolute values ​​of deviation of each abnormal dimension, if the absolute values ​​of deviation of the above-mentioned value clarity dimension and uncertainty dimension are the same, the root cause identification module 16 selects the value clarity dimension as the root cause dimension according to the root cause identification priority (information awareness > value clarity > uncertainty > support > decision effectiveness).

[0090] The meanings and functions of the pre-test and post-test in the pre- and post-test assessment mechanism are as follows: The pre-test is a scale assessment filled out by the patient upon initial login, before knowledge learning and decision-making, used to characterize the patient's initial decision-making state and identify abnormal and root cause dimensions accordingly; the post-test is a scale assessment filled out again by the patient after completing knowledge learning and decision-making, used to characterize the patient's decision-making state after system assistance. The decision assessment module 15 calculates the change in scores between the pre-test and post-test. The change is used to assess the system's intervention effect on the patient's decision-making state. For example, an increase in Prep DM scores and a decrease in DCS scores indicate improved patient decision readiness and reduced decision-making difficulties. The change is included in the decision assistance report by the decision report generation module 20 for medical staff reference, stored by the patient information storage module 11, and used by the statistical analysis module 35 for statistical evaluation of the system's effectiveness.

[0091] 7. System workflow.

[0092] The workflow of the treatment decision support system in this embodiment is as follows: steps S101 to S111.

[0093] In step S101, the patient completes registration and login through the patient terminal 3, fills in basic personal information, and then proceeds to step S102.

[0094] In step S102, the patient completes the pre-test filling of Prep DM and DCS. The decision evaluation module 15 of the platform server 1 calculates the scores of each dimension and compares them with the preset thresholds in the threshold configuration module 14, marks the abnormal dimensions, the root cause identification module 16 determines the root cause dimensions, and then proceeds to step S103.

[0095] In step S103, the personalized push module 19 of the platform server 1 queries the mapping database module 13 based on the root cause dimension and other abnormal dimensions, extracts the corresponding educational modules from the knowledge base storage module 12 in such a way that the proportion of educational modules corresponding to the root cause dimension is not less than 70%, combines them into a personalized learning path and pushes it to the patient terminal 3, and then proceeds to step S104.

[0096] In step S104, the patient browses the content pushed by the learning system through the knowledge learning module 43 of the patient terminal 3. The knowledge quantification and evaluation module 17 records the patient's browsing time, scrolling depth and completion status, and then proceeds to step S105.

[0097] In step S105, the patient reviews the treatment plan, compares the plans, completes the plan preference score and knowledge test according to the standardized decision-making process preset by the decision support module 44, and then proceeds to step S106.

[0098] In step S106, after the patient completes the content of the decision support module 44, they fill in the Prep DM and DCS again as a post-test. The knowledge quantification assessment module 17 determines whether effective learning has been completed and calculates the patient's knowledge acquisition index for each treatment plan. The confidence score calculation module 18 calculates the confidence score based on each treatment plan. Determine the confidence level of the patient's preference rating for the treatment plan, and stratified judgment module 22 combines this with the patient's final choice of treatment plan. Patient preferences are stratified and labeled according to the corresponding confidence level: if And if the confidence level is high, it is marked as a high-confidence preference; if And the confidence level is medium confidence, marked as medium confidence preference; if The confidence level is low, and it is marked as low confidence preference; the hierarchical labeling results are used as the core component of the decision support report, and then proceed to step S107.

[0099] In step S107, the decision report generation module 20 integrates clinical data, scale scores, pre- and post-test changes in scales, confidence levels, protocol preference scores, knowledge test scores, and preference stratification markers from the patient information storage module 11 to generate a decision support report, and then proceeds to step S108.

[0100] In step S108, the decision support report is pushed to the medical terminal 2 through the medical task storage and distribution module 21. Medical staff log in to the system through the medical terminal 2 and view the patient's decision support report and preference hierarchical markings through the patient information viewing module 31, and then proceed to step S109.

[0101] In step S109, medical staff communicate offline with patients and their families based on the decision support report. After jointly determining the treatment plan, the medical staff enters the final treatment plan through the medical terminal 2, and then proceeds to step S110.

[0102] In step S110, the decision report generation module 20 integrates patient clinical information, treatment plan preference scores, patient preferences, and medical opinions to generate a final treatment decision report. The data desensitization module 25 desensitizes the report and patient privacy data, and then sends it to the data server 4 through the platform-side communication module 26. The data is then stored in the desensitized data storage module 53 via the data-side communication module 51 and archived by the classification data management module 54. Then, the process proceeds to step S111.

[0103] In step S111, the patient information storage module 11 stores the patient's clinical data, treatment plan preference scores, medical advice, communication records, and final decision report in a structured manner, and then the process ends.

[0104] It should be noted that the collection, storage, transmission and use of patient-related data involved in the embodiments of this application are all carried out with the user's authorization and comply with the provisions of relevant laws and regulations.

[0105] It is understood that the various numerical designations used in the embodiments of this application are merely for the convenience of description and are not intended to limit the scope of the embodiments of this application.

[0106] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A treatment decision support system for patients with advanced non-small cell lung cancer, characterized in that, include: The system includes a platform server, medical staff terminals, patient terminals, a data server, and a communication network. The platform server communicates with the medical staff terminals, patient terminals, and data server through the communication network. The patient terminal is used to collect basic patient information and relevant data in the decision-making process, display treatment plan information and educational content, guide patients to complete questionnaires, knowledge learning, knowledge tests and treatment plan preference ratings, and upload the collected data to the platform server. The platform server is used to standardize the scores of the decision preparation scale and decision dilemma scale uploaded by the patient terminal, compare them with the preset thresholds, and mark the dimensions with abnormal scores. Based on the labeled abnormal dimensions, a root cause dimension is determined from the abnormal dimensions; based on the abnormal dimension, the dimension-content mapping table is queried, and the educational modules are extracted in the way that the educational modules corresponding to the root cause dimension occupy the dominant number in the overall learning path, combined into a personalized learning path and pushed to the patient terminal. Based on the patient's learning behavior data and knowledge test accuracy in the education module, it is determined whether the patient has completed effective learning, and when it is determined that the patient has completed effective learning, the patient's knowledge acquisition index for each treatment plan is determined. Based on the knowledge acquisition level index, the confidence level and confidence score of patients' preference ratings for each treatment option were determined; Based on the patient's knowledge acquisition index of the final treatment plan, the patient's preferences are stratified and labeled, and a structured treatment decision support report is generated; The medical terminal is used to receive treatment decision support reports and hierarchical labels pushed by the platform server, so that medical staff can view patient information and enter the final treatment plan. The data server is used to store patient data sent by the platform server after it has been anonymized, and to classify, archive, and manage the anonymized data.

2. The treatment decision support system for patients with advanced non-small cell lung cancer according to claim 1, characterized in that, The dimensions for platform server-marked score anomalies include: For each dimension in the decision readiness measure, based on the standardized score of the dimension, if the standardized score is lower than the preset outlier threshold of the decision readiness measure, the dimension is marked as an outlier dimension; for each dimension in the decision dilemma measure, based on the standardized score of the dimension, if the standardized score is higher than the preset outlier threshold of the decision dilemma measure, the dimension is marked as an outlier dimension. Based on the labeled outlier dimensions, the deviation of each outlier dimension is calculated; for the decision readiness metric, the dimension deviation is obtained using the following formula: ； in, This indicates the degree of deviation from the outlier dimensions in the decision readiness measure table; This represents the preset anomaly threshold for the corresponding dimension in the decision readiness metrics table; This represents the standardized actual score of the corresponding dimension in the decision readiness metric. For the decision-making dilemma scale, the dimensional deviation is obtained using the following formula: ； in, Indicates the degree of deviation from the abnormal dimension of the decision-making dilemma scale; This represents the standardized actual score of the corresponding dimension in the Decision Dilemma Scale; This represents the preset abnormal threshold for the corresponding dimension in the Decision Dilemma Scale.

3. The treatment decision support system for patients with advanced non-small cell lung cancer according to claim 2, characterized in that, The platform server determines a root cause dimension from the anomaly dimensions, including: Based on the deviation of each anomaly dimension, the anomaly dimension with the largest absolute value of deviation is selected as the root cause dimension.

4. The treatment decision support system for patients with advanced non-small cell lung cancer according to claim 1, characterized in that, The platform server determines whether a patient has completed effective learning, including: Based on the patient's browsing time, page scrolling depth, module completion status, and knowledge quiz accuracy in each educational module, the patient's effective learning is determined after completing all pushed educational modules and knowledge quizzes. To be considered as having completed effective learning, the following conditions must be met simultaneously: ； in, This indicates the completion coverage rate of the pushed education modules. A completion coverage rate of 100% means that the patient has completed all pushed education modules. Indicates the first The scroll depth of a page with a completed education module; This indicates the patient's total study time; Indicates the first Estimated standard learning time for each educational module; This represents the sum of the estimated standard learning time for all pushed educational modules.

5. The treatment decision support system for patients with advanced non-small cell lung cancer according to claim 4, characterized in that, The platform server determines an index of the patient's knowledge acquisition regarding various treatment options, including: Based on the tags attached to each educational module, the association between the educational module and one or more treatment plans is determined; based on the patient's correct answer rate on knowledge tests for each educational module, and combined with the association between the educational module and each treatment plan, the data is aggregated by treatment plan to obtain the patient's knowledge acquisition index for each treatment plan; where, for the first... For each treatment plan, the patient's knowledge acquisition index of the treatment plan is obtained using the following formula: ； in, Indicates the patient's... The knowledge acquisition index for each treatment plan ranges from 0 to 100. It is a positive integer; Indicates the relationship with the first A collection of educational modules associated with each treatment plan; Represents a set The serial number of the Chinese education module; This represents the patient's accuracy rate on the knowledge test in the i-th educational module, with a value ranging from 0 to 1; Indicates the relationship with the first The number of educational modules associated with each treatment plan.

6. The treatment decision support system for patients with advanced non-small cell lung cancer according to claim 5, characterized in that, The platform server determines the confidence level and confidence score of the patient's preference rating for each treatment option, including: Based on the patient's knowledge acquisition index of each treatment option, the confidence score of the patient's preference rating for each treatment option is obtained; for the first... For each treatment option, the confidence score is obtained using the following formula: ； in, Indicates the patient's... Confidence score of each treatment option preference rating; Indicates the patient's... Knowledge acquisition index of each treatment plan; Based on the A knowledge acquisition index for each treatment option is used to categorize the confidence levels of patients' treatment option preference ratings: when When is a high confidence level; when When is the middle confidence level; when The confidence level is low.

7. The treatment decision support system for patients with advanced non-small cell lung cancer according to claim 1, characterized in that, The platform servers are combined to form a personalized learning path, including: Based on the anomaly dimension, query the dimension-content mapping table to obtain one or more content tags corresponding to each anomaly dimension; based on the content tags, extract the corresponding education modules from the knowledge base, and combine them into a personalized learning path in such a way that the number of education modules corresponding to the root cause dimension accounts for no less than 70% of the overall learning path, push it to the patient's terminal and send a learning reminder notification. The order of multiple educational modules under the same dimension in the personalized learning path is determined by the push priority defined in the dimension-content mapping table.

8. The treatment decision support system for patients with advanced non-small cell lung cancer according to claim 4, characterized in that, The platform server is also used for: When a patient fails to complete effective learning, remind the patient to complete the learning, and only push the educational modules that did not meet the criteria for effective learning to the patient again; Based on the patient's learning behavior data and knowledge test accuracy after completing supplementary learning, the patient's effective learning was reassessed, and the patient's knowledge acquisition index for each treatment plan was re-determined after the assessment was passed.

9. The treatment decision support system for patients with advanced non-small cell lung cancer according to claim 1, characterized in that: The platform server is also used to obtain the pre-test and post-test scores of the Decision Readiness Scale and the Decision Dilemma Scale, and to calculate the change in scores between the two tests. Based on the patient's condition, treatment plan preference score, scale score, changes before and after the test, confidence level, confidence score, preference stratification marker, and medical opinions, a structured treatment decision support report is generated and pushed to the medical terminal linked to the patient, where a task to be processed is generated. Based on the final treatment plan entered into the medical terminal, a final treatment decision report is generated. After anonymizing the final treatment decision report and patient privacy data, the data is sent to the data server for classified archiving and storage.

Images