Speech cognitive impairment assessment and rehabilitation training system

Abstract

The present application belongs to the field of evaluation and rehabilitation technology, and specifically relates to a speech cognitive impairment evaluation and rehabilitation training system, which comprises a double-screen separate control interaction module for splitting the operation authority and interface display content of a therapist control end and a patient display end, receiving patient basic information and task configuration instructions input by a therapist, and independently issuing evaluation tasks and training tasks to the patient display end. A multidimensional evaluation module is used to generate evaluation result data. An adaptive training module is used to output personalized training tasks adapted to the current ability of the patient. A data acquisition and storage module is used to synchronously acquire whole-process traceability data of the patient. A report data export module is used to integrate the evaluation result data, training performance data and whole-process traceability data, generate a comprehensive rehabilitation report, and output a scientific research data set adapted to a scientific research statistical tool; the present application can improve the clinical diagnosis and treatment efficiency and precision, enhance the personalization and effectiveness of rehabilitation training, and expand the clinical application scenarios.

Description

Technical Field

[0001] This invention belongs to the field of assessment and rehabilitation technology, specifically a speech and cognitive impairment assessment and rehabilitation training system. Background Technology

[0002] Speech and cognitive impairment is a common sequela of neurological diseases such as stroke, traumatic brain injury, Alzheimer's disease, and childhood brain development disorders. It mainly manifests as aphasia, impaired comprehension and expression, and cognitive deficits, severely reducing patients' daily communication and social adaptation abilities. Currently, electronic assessment and training systems are widely used in clinical rehabilitation, replacing traditional paper-based questionnaires and one-on-one manual training, thus initially achieving standardization of the rehabilitation process and improving diagnostic and treatment efficiency.

[0003] However, existing technologies still have many inherent shortcomings: First, the systems generally adopt a single-screen integrated display architecture, and the therapist's operating interface and the patient's training interface are not separated in terms of permissions and content. Irrelevant content such as backend settings and function menus can easily distract patients with attention deficit, causing distraction during training and accidental touches. Simultaneously, therapists cannot independently coordinate the entire rehabilitation process, resulting in insufficient clinical management efficiency and inaccurate assessment results. Second, the assessment dimensions and training methods are limited, with assessments often relying on simple screening scales and lacking a standardized, comprehensive speech and cognitive assessment system. This makes it impossible to accurately classify and grade impairments, and difficult to adapt to the progressive rehabilitation needs of patients with different degrees of impairment. Third, the systems are mostly closed and fixed architectures, with fixed training question banks and rehabilitation plans from the factory, not supporting custom expansion by clinicians. This leads to poor adaptability to special cases, difficulty in quickly implementing cutting-edge rehabilitation technologies, and high costs for subsequent equipment upgrades.

[0004] In summary, existing technologies cannot meet the refined and individualized clinical needs of modern rehabilitation, which restricts the overall quality and clinical application value of speech and cognitive disorder rehabilitation. There is an urgent need for a speech and cognitive disorder assessment and rehabilitation training system to solve the above-mentioned deficiencies. Summary of the Invention

[0005] To overcome the shortcomings of existing technologies, this invention proposes a speech and cognitive impairment assessment and rehabilitation training system. This invention primarily addresses the problems of existing systems, such as confusing operation permissions and interface interference, limited assessment dimensions and training methods, and rigid rehabilitation plans.

[0006] The technical solution adopted by this invention to solve its technical problem is: a speech and cognitive disorder assessment and rehabilitation training system provided by this invention, comprising:

[0007] The dual-screen control and interaction module is used to separate the operation permissions and interface display content of the therapist's control terminal and the patient's display terminal. It receives the patient's basic information and task configuration instructions entered by the therapist, and independently issues assessment tasks and training tasks to the patient's display terminal.

[0008] The multi-dimensional assessment module is used to call the localized Chinese test question bank to perform assessment tasks, collect patient response data, and generate assessment result data through multi-dimensional quantitative analysis algorithms.

[0009] The adaptive training module is used to match the initial rehabilitation training plan based on the assessment results data. During the execution of the training task, it reads the patient's training performance data in real time, dynamically adjusts the training difficulty gradient and guidance level through the difficulty adaptive algorithm, and outputs personalized training tasks that are adapted to the patient's current ability.

[0010] The data acquisition and storage module is used to simultaneously collect patients' responses, reaction times, voice data, writing trajectories, and interactive operation data during the evaluation and training process, generating full-process traceability data.

[0011] The report data export module is used to integrate assessment result data, training performance data, and full-process traceability data to generate a comprehensive rehabilitation report and output a research dataset adapted to scientific research statistical tools.

[0012] The present invention provides a speech and cognitive disorder assessment and rehabilitation training system, the steps of separating operation permissions and interface display content include:

[0013] The industrial control host is equipped with two independent screen output channels, which are respectively connected to the therapist's control screen and the patient's screen.

[0014] The therapist's control screen loads a global operation interface, displaying a list of patient medical records, overall control of assessment and training, parameter settings, protocol editing, and data analysis functions. It also receives basic patient information and task configuration instructions input by the therapist.

[0015] The patient screen loads a simplified interactive interface, displaying only the currently performed assessment or training task.

[0016] Maintain real-time data synchronization between the therapist's control terminal and the patient's display terminal, synchronously send the therapist's configuration commands to the patient's terminal, and synchronously transmit the patient's response data back to the therapist's terminal.

[0017] The present invention provides a speech and cognitive impairment assessment and rehabilitation training system, the steps of which include generating assessment result data are as follows:

[0018] Based on the task configuration instructions, the localized Chinese question bank for the corresponding assessment dimension is invoked, the assessment process is generated, and pushed to the patient's display terminal.

[0019] Collect patient responses, correct / incorrect answers, and response times during the assessment process to generate patient response data.

[0020] Based on standardized scoring rules, and combined with algorithms for calculating aphasia quotient and cortical quotient, the patient's response data was quantified to obtain functional quantitative indicators.

[0021] Based on the functional quantification indicators and the aphasia classification rules, the disorder type and severity level are determined, and a sub-item score map is generated.

[0022] Retrieve the patient's historical assessment data and compare it longitudinally with the current assessment results to generate dynamic trend analysis data.

[0023] By integrating functional quantitative indicators, sub-item score maps, and dynamic trend analysis data, the evaluation results data are obtained.

[0024] The present invention provides a speech and cognitive impairment assessment and rehabilitation training system, the steps of which include dynamically adjusting the training difficulty gradient and guidance level as follows:

[0025] A smooth difficulty gradient of 9-40 levels and a guidance logic of ≥99 levels are preset as the basic level for difficulty adjustment.

[0026] The system collects data in real time on the patient's accuracy, reaction time, number of prompts for help, and error type for each round of training, generating training performance data.

[0027] When the accuracy rate reaches a preset threshold and the reaction time is stable within a reasonable range after multiple rounds of training, the current training difficulty gradient is automatically increased and the guidance level is reduced.

[0028] When the training error rate exceeds the preset threshold, the reaction time is significantly prolonged, or the prompt for help is triggered multiple times, the current training difficulty gradient is automatically lowered and more guidance steps are added.

[0029] Based on the adjusted difficulty gradient and guidance level, the output parameters of subsequent training tasks are updated.

[0030] This invention provides a speech and cognitive impairment assessment and rehabilitation training system, whose adaptive training module also supports the construction of a custom training library. The steps include:

[0031] It features an open, visual editing interface, allowing therapists to upload training materials in the form of images, audio, or text.

[0032] It receives the question rules, difficulty parameters, training duration and feedback mode customized by the therapist, and generates the training task logic.

[0033] The training materials are bound and encapsulated with the training task logic to generate a personalized training library.

[0034] When matching the initial rehabilitation training plan, the personalized training library is included in the pool of optional training plans, which can be directly invoked and executed.

[0035] The present invention provides a speech and cognitive impairment assessment and rehabilitation training system, the steps of generating full-process traceability data include:

[0036] Throughout the assessment and training process, the system simultaneously captures patients' touch clicks and drag-and-drop connection operation data, microphone-collected voice recording data, handwriting input trajectory data, and the answer results and reaction time for each question, forming multi-source data according to the preset collection frequency.

[0037] Timestamp alignment is performed on multi-source data to form a raw process dataset with a single task dimension.

[0038] The original process dataset is classified and archived according to four dimensions: unique patient medical record identifier, assessment / training program, execution date, and rehabilitation stage, to generate full-process traceability data.

[0039] This invention provides a speech and cognitive impairment assessment and rehabilitation training system. The localized Chinese assessment question bank includes a speech assessment question bank and a cognitive assessment question bank, wherein the assessment dimensions of the cognitive assessment question bank include:

[0040] The assessment includes eight neurocognitive dimensions: visuospatial construction, executive transformation, naming memory, attentional language, verbal fluency, abstract thinking, delayed recall, and spatiotemporal orientation.

[0041] When performing cognitive assessment tasks, patient response data for the corresponding dimensions are collected, scores for each dimension are calculated, cognitive function sub-assessment results are generated, and the results are incorporated into the assessment data.

[0042] This invention provides a speech and cognitive impairment assessment and rehabilitation training system. The adaptive training module also supports multimodal interaction and adaptation to specific conditions. The steps include:

[0043] It provides independent on / off switches and combination configuration options for visual and auditory feedback, allowing users to select the appropriate feedback mode based on their visual and auditory impairments.

[0044] For patients with unilateral neglect, the training information bar can be freely switched between the left and right sides of the interface to adapt to the patient's visual field defect.

[0045] It offers customizable adjustment functions for training background color temperature and interface layout to match the needs of cognitive psychology rehabilitation environments.

[0046] It has three built-in training systems for adults, the elderly and children, respectively corresponding to general speech and cognitive rehabilitation, sensory stimulation rehabilitation for the elderly and attention training for children. The corresponding training system is matched according to the patient's age and symptoms.

[0047] The present invention provides a speech and cognitive impairment assessment and rehabilitation training system, the steps of which include generating a comprehensive rehabilitation report are as follows:

[0048] Retrieve the corresponding patient's assessment results data, phased training performance data, and a summary of the entire process traceability data.

[0049] Data from multiple modules and time periods are integrated and summarized, and the report content is organized according to the structure of patient basic information, assessment scores, training programs, efficacy trends, and therapist recommendations.

[0050] It supports importing custom report templates, allowing you to match and fill the integrated report content into the corresponding positions in the template to generate a comprehensive rehabilitation report.

[0051] The present invention provides a speech and cognitive impairment assessment and rehabilitation training system, the steps of which include outputting a research dataset are as follows:

[0052] Using the patient's unique medical record identifier as an index, longitudinal data on the patient's historical assessments and training are extracted.

[0053] The longitudinal data is grouped and organized according to the time dimension, and the data field format and statistical caliber are standardized.

[0054] Generates research datasets in a preset target format. The dataset fields are adapted to statistical analysis software, and it supports batch extraction of research data from multiple patients to form research datasets.

[0055] The beneficial effects of this invention are as follows:

[0056] 1. This invention utilizes a dual-screen control architecture, allowing therapists to complete all settings and monitoring in one stop at the control terminal, while protecting patients from irrelevant interference. This completely solves the problems of distraction and accidental operation during training for patients with cognitive impairment, effectively improving the efficiency of rehabilitation training and the accuracy of assessments. Simultaneously, it simplifies the therapist's operational procedures and improves the efficiency of single-station diagnosis and treatment. Through the assessment system, it achieves refined, visualized, and dynamic diagnosis and localization of speech and cognitive impairments. The output of multi-dimensional quantitative indicators solves the problems of one-sided assessments and inaccurate classifications in traditional equipment, ensuring that assessment results fully conform to clinical diagnosis and treatment standards, significantly improving the accuracy and professionalism of disease diagnosis and localization. The collection and archiving of traceable data throughout the entire process allows for review and retrospection of the training process at any time, providing original data support for clinical case demonstrations and advanced medical research.

[0057] 2. The proprietary training system in this invention covers patients of all ages, and features such as free combination of multimodal feedback and adaptive switching of the unilateral neglect interface enable patients with special conditions to effectively participate in rehabilitation training, greatly expanding the clinical application scenarios of the device. Intelligent adaptive training eliminates the need for therapists to manually monitor and adjust parameters throughout the rehabilitation process, significantly reducing repetitive procedures for medical staff and lowering the time and effort required. Attached Figure Description

[0058] The invention will now be further described with reference to the accompanying drawings.

[0059] Figure 1This is a schematic diagram of a module of a speech and cognitive impairment assessment and rehabilitation training system provided in an embodiment of the present invention;

[0060] Figure 2 yes Figure 1 A flowchart illustrating the process by which the multi-dimensional evaluation module obtains evaluation result data.

[0061] Figure 3 yes Figure 1 A flowchart illustrating the process of generating a comprehensive rehabilitation report using the data export module. Detailed Implementation

[0062] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0063] like Figures 1 to 3 As shown in the figure, an embodiment of the present invention provides a speech and cognitive impairment assessment and rehabilitation training system, the system comprising:

[0064] The dual-screen control and interaction module is used to separate the operation permissions and interface display content of the therapist's control terminal and the patient's display terminal. It receives the patient's basic information and task configuration instructions entered by the therapist, and independently issues assessment tasks and training tasks to the patient's display terminal.

[0065] The steps to separate operation permissions from interface display content include:

[0066] The industrial control host is equipped with two independent screen output channels, which are respectively connected to the therapist's control screen and the patient's screen.

[0067] The hardware platform is a mobile medical cart (mobile cabinet), with a total platform length of 1400mm and a platform height of 780mm from the ground. The platform edges are decorated with light green rounded corner trim with a radius of R30mm. Four swivel casters with brakes are installed at the bottom of the cart, allowing for flexible movement between wards and examination rooms. The right-side enclosed compartment houses an industrial control unit equipped with a 9th-generation i7 processor, 16GB of RAM, and a 1TB SSD. The SSD stores all assessment scales, training materials, and patient training data throughout the entire process. The industrial control unit features dual independent video output interfaces (HDMI + DP), corresponding to two independent rendering channels of the GPU, with physical isolation between the video memory and rendering resources. Two 21.5-inch LCD monitors are arranged in parallel on the cart's platform: the left one is the therapist's control screen, and the right one is the patient's screen. Each monitor connects to one of the industrial control unit's two independent video output channels, ensuring independent output and no interference.

[0068] The therapist's control screen loads a global operation interface, displaying a list of patient medical records, overall control of assessment and training, parameter settings, protocol editing, and data analysis functions. It also receives basic patient information and task configuration instructions input by the therapist.

[0069] Medical Record Management Area: Supports creating new patient files, allowing users to enter basic information such as name, gender, age, medical record number, diagnosis, disease course, and disability type. Each patient is assigned a unique medical record identifier, which serves as the primary key for data association across the entire system. Supports querying historical records and batch management of patient data.

[0070] Task Control Area: Visually select verbal assessment, cognitive assessment, comprehensive assessment, and corresponding training items, and send tasks to the patient with one click. Real-time monitoring of task progress and patient response status, supporting pause, termination, and mid-task adjustments.

[0071] Parameter configuration area: Set training duration, upper and lower limits of difficulty, feedback mode, interface layout, and parameters for special diseases. All parameter modifications are synchronized to the training engine in real time.

[0072] Program editing area: Connects to the open training library creation platform, supporting the creation, editing, importing and exporting of personalized training programs.

[0073] Data Analysis Area: Visually displays patient recovery trends, radar charts of scores across various dimensions, training completion rate statistics, and other charts.

[0074] The patient screen loads a simplified interactive interface, displaying only the currently performed assessment or training task.

[0075] All UI elements are marked with permission tags, and only elements marked as visible to patients will be rendered and displayed, thus blocking irrelevant content such as background menus, settings entry points, and other patient information from the source.

[0076] The patient-side interface retains only three core areas: the task display area (question content and prompts), the interactive control area (selection button, handwriting area, and voice input button), and the feedback prompt area (correction feedback and encouragement prompts), with no entry point to the backend.

[0077] The interface elements are designed with large sizes and high contrast to reduce cognitive load. For patients with attention deficit disorder, all pop-ups and animated distractions are blocked to ensure focus on tasks.

[0078] Maintain real-time data synchronization between the therapist's control terminal and the patient's display terminal, synchronously send the therapist's configuration commands to the patient's terminal, and synchronously transmit the patient's response data back to the therapist's terminal.

[0079] The formula for calculating synchronization delay is expressed as follows:

[0080]

[0081] In the formula, It is the average synchronization delay. It is the timestamp of the instruction / data sender. It is the timestamp of the instruction / data receiving end. It is the total number of synchronous samples. It is the first Secondary transmission sample.

[0082] The multi-dimensional assessment module is used to call the localized Chinese test question bank to perform assessment tasks, collect patient response data, and generate assessment result data through multi-dimensional quantitative analysis algorithms.

[0083] The steps to generate evaluation result data include:

[0084] Based on the task configuration instructions, the localized Chinese question bank for the corresponding assessment dimension is invoked, the assessment process is generated, and pushed to the patient's display terminal.

[0085] Verbal assessment question bank: Fully covers all core clinical assessment items, including spontaneous speech, listening comprehension, retelling, naming, reading, writing, application, structure, noun comprehension, verb comprehension, sentence comprehension, execution of oral commands, cartoon description and picture description, with a total of ≥1200 questions.

[0086] Cognitive assessment question bank: It completes a comprehensive assessment of neurocognition, visual perception and verbal psychology from eight dimensions: visuospatial construction, executive conversion, naming memory, attentional language, word fluency, abstract thinking, delayed recall and spatiotemporal orientation, with a total of ≥800 questions.

[0087] Collect patient responses, correct / incorrect answers, and response times during the assessment process to generate patient response data.

[0088] Patients can interact and answer questions through various methods, including touch clicking, dragging, connecting lines, voice responses, handwriting input, and answer selection. A gooseneck microphone is installed in front of the patient screen and connected to the industrial control host sound card to capture the patient's voice responses.

[0089] The formula for calculating reaction time is expressed as follows:

[0090]

[0091] In the formula, It refers to the reaction time for a single question. It is the timestamp when the question rendering is complete and the timer starts. It is the timestamp of the patient clicking submit and completing the answer.

[0092] Based on standardized scoring rules, and combined with algorithms for calculating aphasia quotient and cortical quotient, the patient's response data was quantified to obtain functional quantitative indicators.

[0093] Aphasia Quotient (AQ) is used to quantify a patient's overall speech function level and is a core indicator for assessing the severity of aphasia. The calculation formula is as follows:

[0094]

[0095] In the formula, It is the aphasia quotient. It is the total number of verbal assessment sub-items. No. The actual score of each item It is the first The full score for each sub-item. No. The weighting coefficients for each sub-item are set based on clinical norms, such as 0.22 for auditory comprehension, 0.18 for repetition, 0.22 for naming, 0.14 for reading, 0.14 for writing, and 0.10 for spontaneous speech. A higher weight indicates a greater impact of that dimension on overall speech function.

[0096] The Cortical Quotient (CQ) incorporates a nonverbal cognitive dimension into the verbal function, reflecting the overall functional level of the cerebral cortex. Its calculation formula is as follows:

[0097]

[0098] In the formula, It is a leather merchant. It is the total number of sub-items in the comprehensive evaluation. No. The actual score of each item It is the first The full score for each sub-item. No. The weight coefficient of each sub-item.

[0099] Based on the functional quantification indicators and the aphasia classification rules, the disorder type and severity level are determined, and a sub-item score map is generated.

[0100] Based on the relative patterns of scores for each sub-item, and matching with classic clinical aphasia classification features, the system automatically provides classification suggestions. Core rule table:

[0101] Broca's aphasia (motor aphasia) The scores for expression were significantly lower than those for comprehension, indicating a lack of fluency in speech. Broca's area of ​​the left inferior frontal gyrus Wernicke aphasia (sensory aphasia) The comprehension score was significantly lower than the expression score; the speech was fluent but empty. Wernicke's area in the left superior temporal gyrus Conductive aphasia The retelling score was significantly lower than the listening comprehension and naming scores. Left arcuate fasciculus injury Complete aphasia All verbal sub-items scores decreased significantly. Extensive damage to the left speech area Naming aphasia The naming score was significantly low, while the other dimensions were relatively well-retained. Left temporo-occipital junction

[0102] Severity grading: Based on the Boston Aphasia Severity Scale, the impairment is divided into 6 levels by combining the aphasia quotient score.

[0103] Two major visualization charts are generated: a major category quantile chart (a radar chart of scores in the first-level dimensions of verbal / cognitive speech) and a minor category quantile chart (a bar chart of scores for each sub-item), which intuitively show the patient's functional weaknesses and strengths.

[0104] Retrieve the patient's historical assessment data and compare it longitudinally with the current assessment results to generate dynamic trend analysis data.

[0105] The formula for calculating the improvement rate of therapeutic effect is expressed as follows:

[0106]

[0107] In the formula, It is the improvement rate of aphasia quotient. This is the aphasia quotient score in this assessment. It is the aphasia quotient score from the last assessment.

[0108] A line graph of rehabilitation trends was drawn with assessment time on the horizontal axis and AQ / CQ scores on the vertical axis, marking the training programs at each stage to help therapists judge the effectiveness of the programs.

[0109] By integrating functional quantitative indicators, sub-item score maps, and dynamic trend analysis data, the evaluation results data are obtained.

[0110] The localized Chinese assessment question bank includes a verbal assessment question bank and a cognitive assessment question bank. The cognitive assessment question bank includes the following assessment dimensions:

[0111] The assessment includes eight neurocognitive dimensions: visuospatial construction, executive transformation, naming memory, attentional language, verbal fluency, abstract thinking, delayed recall, and spatiotemporal orientation.

[0112] Neurocognitive Assessment Scale:

[0113] Visual space construction Spatial perception and visual construction ability Block pattern copying and graphic assembly Perform conversion Cognitive flexibility, task switching ability Connection quiz, color word interference task Naming memory Semantic memory and vocabulary retrieval ability Immediate recall and delayed recall of words Pay attention to language Selective attention, verbal attention Digital breadth, auditory target detection Fluent in word choice Speech generation and semantic association abilities Category word fluency test Abstract thinking Logical reasoning and conceptual generalization abilities Similarity tests, classification tasks Delayed recall Long-term memory, information retention ability Delayed recall of words and recognition of images Spatiotemporal orientation Temporal and spatial orientation capabilities Date and location orientation, scene orientation determination

[0114] When performing cognitive assessment tasks, patient response data for the corresponding dimensions are collected, scores for each dimension are calculated, cognitive function sub-assessment results are generated, and the results are incorporated into the assessment data.

[0115] The adaptive training module is used to match the initial rehabilitation training plan based on the assessment results data. During the execution of the training task, it reads the patient's training performance data in real time, dynamically adjusts the training difficulty gradient and guidance level through the difficulty adaptive algorithm, and outputs personalized training tasks that are adapted to the patient's current ability.

[0116] The steps for dynamically adjusting the training difficulty gradient and guidance level include:

[0117] A smooth difficulty gradient of 9-40 levels and a guidance logic of ≥99 levels are preset as the basic level for difficulty adjustment.

[0118] Smooth difficulty gradient design: Difficulty is determined by five parameters: vocabulary complexity, sentence length, number of distractors, semantic relevance, and time limit. Only one or two parameters are adjusted between adjacent difficulty levels to ensure a smooth transition in difficulty and avoid significant frustration or stagnation in the comfort zone for patients.

[0119] Multi-layered guidance logic design: ≥99 guidance levels are set, with guidance intensity increasing progressively from no prompts to direct answer display. Guidance methods include semantic prompts, initial sound prompts, lip-reading prompts, whole word prompts, and step-by-step prompts. Each level corresponds to a different combination of prompt intensity to ensure that patients of varying degrees receive appropriate assistance.

[0120] The system collects data in real time on the patient's accuracy, reaction time, number of prompts for help, and error type for each round of training, generating training performance data.

[0121] Accuracy: Number of correct answers in the current round / Total number of questions, with a value range of 0 to 1.

[0122] Reaction time: Average reaction time per question, calculated using logic completely consistent with the evaluation module.

[0123] Number of prompts for help: The total number of times the patient actively triggered prompts, reflecting the challenge of the task to the patient.

[0124] Error types: Incorrect answers are categorized into four types: semantic errors, phonetic errors, syntactic errors, and operational errors. This is used to locate the damaged areas and help adjust the training focus.

[0125] When the accuracy rate reaches a preset threshold and the reaction time is stable within a reasonable range after multiple rounds of training, the current training difficulty gradient is automatically increased and the guidance level is reduced.

[0126] When the training error rate exceeds the preset threshold, the reaction time is significantly prolonged, or the prompt for help is triggered multiple times, the current training difficulty gradient is automatically lowered and more guidance steps are added.

[0127] Based on the adjusted difficulty gradient and guidance level, the output parameters of subsequent training tasks are updated.

[0128] An adaptive algorithm that uses a weighted comprehensive performance score plus consecutive rounds of judgment is adopted to achieve smooth adjustment of difficulty, adjusting only one difficulty level at a time to avoid abrupt changes in difficulty.

[0129] The formula for calculating the overall performance score is as follows:

[0130]

[0131] In the formula, It is a score based on overall performance in a single round. It is the accuracy rate in the current round. It is the standardized reaction time, calculated using the following formula: The faster the reaction, the higher the score. This represents the average reaction time of the current round. This is the shortest reference time for this difficulty level. This is the longest reference time for this difficulty level. It is the standardized number of prompts, calculated using the following formula: The fewer prompts required, the higher the score. This represents the number of times a request for help has been prompted in the current round. This represents the maximum number of hints allowed per round at this difficulty level. It is the weighting coefficient.

[0132] Promotion / relegation determination rules:

[0133] Upgrade requirements: 3 consecutive rounds →Difficulty level +1, guidance level -1;

[0134] Relegation conditions: 2 consecutive rounds →Difficulty level -1, guidance level +2;

[0135] Stable conditions: In all other cases, the current difficulty and guidance level remain unchanged.

[0136] The adaptive training module also supports building custom training libraries, the steps of which include:

[0137] It features an open, visual editing interface, allowing therapists to upload training materials in the form of images, audio, or text.

[0138] It receives the question rules, difficulty parameters, training duration and feedback mode customized by the therapist, and generates the training task logic.

[0139] The training materials are bound and encapsulated with the training task logic to generate a personalized training library.

[0140] When matching the initial rehabilitation training plan, the personalized training library is included in the pool of optional training plans, which can be directly invoked and executed.

[0141] Material Upload: The system features an open visual editing interface that allows therapists to batch upload images (JPG / PNG), audio (MP3 / WAV), and text materials. The system automatically standardizes the formats and compresses the data for storage.

[0142] Rule configuration: Through drag-and-drop visual configuration, therapists can customize question stems, correct answers, distractor options, difficulty labels, feedback content, and judgment rules without any programming background.

[0143] Encapsulation and storage: The materials and rules are bound and encapsulated into independent training packages to generate a local personalized training library, which supports import and export and management by disease type.

[0144] Solution retrieval: When matching training solutions, the personalized training library and the system's built-in question bank are both included in the pool of available solutions, which therapists can directly select and use.

[0145] The adaptive training module also supports multimodal interaction and adaptation to specific conditions, including the following steps:

[0146] It provides independent on / off switches and combination configuration options for visual and auditory feedback, allowing users to select the appropriate feedback mode based on their visual and auditory impairments.

[0147] For patients with unilateral neglect, the training information bar can be freely switched between the left and right sides of the interface to adapt to the patient's visual field defect.

[0148] It offers customizable adjustment functions for training background color temperature and interface layout to match the needs of cognitive psychology rehabilitation environments.

[0149] It has three built-in training systems for adults, the elderly and children, respectively corresponding to general speech and cognitive rehabilitation, sensory stimulation rehabilitation for the elderly and attention training for children. The corresponding training system is matched according to the patient's age and symptoms.

[0150] Feedback mode configuration: Provides independent switches for visual and auditory feedback, supporting individual or combined activation, suitable for patients with visual and auditory impairments.

[0151] Unilateral neglect adaptation: Supports one-click switching of the training information bar on the left / right side of the interface, shifting the core interaction area towards the patient's healthy side. The shift ratio can be customized (10%~30%), adapting to the visual field characteristics of patients with unilateral spatial neglect.

[0152] Environmental parameter adjustment: Supports customization of training background color temperature, font size and interface layout to match the optimal rehabilitation environment in cognitive psychology and alleviate patients' visual fatigue and inattention.

[0153] All-age training system: Includes three exclusive training systems for adults, seniors, and children.

[0154] Adult System: Focuses on speech function recovery and daily communication, with a simple and professional interface, including training in workplace and daily life scenarios.

[0155] The elderly system focuses on cognitive maintenance and sensory stimulation, using large fonts and high-contrast design, and includes sensory stimulation (sight, hearing, touch, smell, and taste) and nostalgic cognitive training.

[0156] Children's System: Focuses on developmental improvement and concentration training, using cartoonish and fun designs, including gamified concentration and language development training.

[0157] The data acquisition and storage module is used to simultaneously collect patients' responses, reaction times, voice data, writing trajectories, and interactive operation data during the evaluation and training process, generating full-process traceability data.

[0158] The steps to generate full-process traceability data include:

[0159] Throughout the assessment and training process, the system simultaneously captures patients' touch clicks and drag-and-drop connection operation data, microphone-collected voice recording data, handwriting input trajectory data, and the answer results and reaction time for each question, forming multi-source data according to the preset collection frequency.

[0160] Multi-source data tables:

[0161] 1 Touch operation data Coordinate point sequence 60Hz Touchscreen driver Analyze operational fluency and attention trajectory 2 Drag and drop data Path coordinate sequence 30Hz Interaction Engine Analyzing visual-motor coordination ability 3 Voice recording data Audio stream (WAV) 16kHz Gooseneck microphone + sound card Voice analysis, process retrospection, case demonstration 4 Writing trajectory data Handwriting coordinates + pressure 120Hz Handwriting screen driver Analyze writing fluency and character formation ability 5 Answer result data Structured text One question per question Answer Judgment Module Accuracy statistics and difficulty adaptation 6 Reaction time data Value (seconds) One question per question Timing module Processing smoothness analysis

[0162] Multi-source data is timestamped to form a raw process dataset with a single task dimension. Millisecond-level time alignment is achieved using a unified clock reference and offset calibration, ensuring that all modal data at the same moment correspond one-to-one, supporting frame-by-frame playback of the patient's complete response process.

[0163] The original process dataset is classified and archived according to four dimensions: unique patient medical record identifier, assessment / training program, execution date, and rehabilitation stage, to generate full-process traceability data.

[0164] A four-level directory archiving architecture is adopted to store the entire process data in a structured manner:

[0165] First-level directory: A unique medical record identifier for each patient, enabling physical isolation of patient data.

[0166] Second-level category: Evaluation projects / training projects, distinguishing different types of task data.

[0167] Third-level directory: Execution date, archives single task data by day.

[0168] Level 4 directory: Rehabilitation phase (acute phase / recovery phase / maintenance phase), facilitating phased statistical analysis.

[0169] The report data export module is used to integrate assessment result data, training performance data, and full-process traceability data to generate a comprehensive rehabilitation report and output a research dataset adapted to scientific research statistical tools.

[0170] The steps to generate a comprehensive rehabilitation report include:

[0171] Retrieve the corresponding patient's assessment results data, phased training performance data, and a summary of the entire process traceability data.

[0172] Data from multiple modules and time periods are integrated and summarized, and the report content is organized according to the structure of patient basic information, assessment scores, training programs, efficacy trends, and therapist recommendations.

[0173] The report adopts a standardized seven-section structure, covering all the information needed for clinical diagnosis and treatment: patient basic information page, speech and cognitive assessment results, phase training summary, rehabilitation efficacy trend, core shortcomings and rehabilitation suggestions, next phase training plan and therapist signature column.

[0174] It supports importing custom report templates, allowing you to match and fill the integrated report content into the corresponding positions in the template to generate a comprehensive rehabilitation report.

[0175] Medical institutions can upload their own standardized Word report templates and set placeholders in the corresponding locations. The integrated data and charts are then matched and filled into the corresponding placeholders to generate a report that conforms to hospital documentation standards. Exporting to Word / PDF format is supported, and the report can be directly connected to the cart's built-in printer for paper report output.

[0176] The steps to output a research dataset include:

[0177] Using the patient's unique medical record identifier as an index, longitudinal data on the patient's historical assessments and training are extracted.

[0178] The longitudinal data is grouped and organized according to the time dimension, and the data field format and statistical caliber are standardized.

[0179] Generates research datasets in a preset target format. The dataset fields are adapted to statistical analysis software, and it supports batch extraction of research data from multiple patients to form research datasets.

[0180] Dataset format: standard Excel two-dimensional table format, the first row is the variable name, and each row corresponds to one evaluation / training record. It is fully compatible with the import requirements of statistical tools such as SPSS and R, and no manual secondary processing is required.

[0181] Examples of core fields: medical record number, gender, age, diagnosis, course of illness, assessment date, aphasia quotient, cortical quotient, scores of each sub-item, total training time, average accuracy and improvement rate, etc.

[0182] Batch export: Supports filtering patient groups by diagnosis type, age group, rehabilitation plan and other conditions, and batch exporting large sample datasets to support clinical research and big data analysis.

[0183] Example of a complete clinical workflow:

[0184] Step 1: The therapist creates a new patient medical record on the therapist control panel and enters basic information;

[0185] Step 2: The therapist selects the standardized aphasia assessment on the therapist control panel and issues the assessment task; the assessment instructions are only displayed synchronously on the patient's display panel.

[0186] Step 3: The patient faces the patient display terminal to complete all assessments, and the microphone simultaneously captures the patient's spoken speech.

[0187] Step 4: After the assessment, the system automatically calculates the aphasia quotient, cortical quotient, and generates a classification and comparison chart;

[0188] Step 5: Based on the assessment results, the therapist selects the factory-preset training program or enters the open training library platform to create a personalized training library; sets the training difficulty range, training duration, and feedback mode; if the patient has right-sided neglect, the training information bar is set to the left side of the screen;

[0189] Step 6: Start training. The patient's display only shows the training content and analyzes the answer results in real time, automatically and smoothly adjusting the difficulty level; the system records the original data of writing, connecting lines and speech throughout the process.

[0190] Step 7: After the rehabilitation cycle is completed, the reporting unit integrates all language and cognitive assessment data from multiple assessments and uses a custom Word template to output a comprehensive report; the research export unit exports all longitudinal data of the medical record as an Excel file, which can be directly imported into SPSS.

[0191] In summary, the speech and cognitive disorder assessment and rehabilitation training system provided in this embodiment significantly improves therapists' management efficiency and diagnostic professionalism through dual-screen control and a comprehensive assessment system, enabling more precise rehabilitation plan development and enhancing clinical efficacy and accuracy. Multi-level training content and intelligent adaptive adjustment ensure precise matching between the training plan and the patient's abilities, fundamentally improving the effectiveness of rehabilitation training and enhancing its personalization and effectiveness. Intelligent difficulty adaptation and automatic report generation reduce therapists' repetitive parameter adjustments and paperwork, allowing them to focus more on core rehabilitation guidance, simplifying manual operations and reducing energy consumption. Adaptability to specific conditions and all age groups greatly broadens the device's applicability, enabling it to serve a wider range of patients.

[0192] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0193] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A speech and cognitive impairment assessment and rehabilitation training system, characterized in that, include: The dual-screen control and interaction module is used to separate the operation permissions and interface display content of the therapist control terminal and the patient display terminal. It receives the patient's basic information and task configuration instructions entered by the therapist and independently issues assessment tasks and training tasks to the patient display terminal. The multi-dimensional assessment module is used to call the localized Chinese test question bank to perform the assessment task, collect patient response data, and generate assessment result data through a multi-dimensional quantitative analysis algorithm. The adaptive training module is used to match an initial rehabilitation training plan based on the assessment result data, read the patient's training performance data in real time during the execution of the training task, dynamically adjust the training difficulty gradient and guidance level through the difficulty adaptive algorithm, and output a personalized training task that is adapted to the patient's current ability. The data acquisition and storage module is used to simultaneously collect patients' response results, reaction time, voice data, writing trajectory and interactive operation data during the evaluation and training process, and generate full-process traceability data; The report data export module is used to integrate the assessment result data, training performance data, and the whole process traceability data to generate a comprehensive rehabilitation report and output a research dataset adapted to scientific research statistical tools.

2. The speech and cognitive impairment assessment and rehabilitation training system according to claim 1, characterized in that: The steps to separate operation permissions from interface display content include: The industrial control host is equipped with two independent screen output channels, which are respectively connected to the therapist's control screen and the patient's screen; The therapist control screen loads a global operation interface, displaying a patient medical record list, assessment and training control, parameter settings, protocol editing and data analysis functions, and receiving basic patient information and task configuration instructions input by the therapist. The patient screen loads a simplified interactive interface, displaying only the currently performed assessment or training task; Maintain real-time data synchronization between the therapist's control terminal and the patient's display terminal, synchronously send the therapist's configuration commands to the patient's terminal, and synchronously transmit the patient's response data back to the therapist's terminal.

3. The speech and cognitive impairment assessment and rehabilitation training system according to claim 1, characterized in that: The steps for generating the evaluation result data include: The task configuration instructions call the localized Chinese question bank for the corresponding assessment dimension, generate the assessment process, and push it to the patient's display terminal. Collect the patient's responses, correct / incorrect answers, and reaction time during the assessment process to form patient response data; Based on standardized scoring rules, and combined with aphasia quotient and cortical quotient calculation algorithms, the patient's response data were quantitatively calculated to obtain functional quantitative indicators; Based on the functional quantification indicators, match the aphasia classification rules to determine the type and severity of the disorder, and generate a sub-item score map. Retrieve the patient's historical assessment data and compare it longitudinally with the results of this assessment to generate dynamic trend analysis data; The evaluation result data is obtained by integrating the functional quantitative indicators, the sub-item score map, and the dynamic trend analysis data.

4. The speech and cognitive impairment assessment and rehabilitation training system according to claim 1, characterized in that: The steps for dynamically adjusting the training difficulty gradient and guidance level include: A smooth difficulty gradient of 9-40 levels and a guidance logic of ≥99 layers are preset as the basic level for difficulty adjustment; The system collects data in real time on the patient's accuracy, reaction time, number of prompts for help, and error type for each round of training, generating training performance data. When the accuracy rate reaches a preset threshold and the reaction time is stable within a reasonable range after multiple rounds of training, the current training difficulty gradient is automatically increased and the guidance level is reduced. When the training error rate exceeds the preset threshold, the reaction time is significantly prolonged, or the prompt for help is triggered multiple times, the current training difficulty gradient is automatically lowered and more guidance steps are added. Based on the adjusted difficulty gradient and guidance level, the output parameters of subsequent training tasks are updated.

5. The speech and cognitive impairment assessment and rehabilitation training system according to claim 1, characterized in that: The adaptive training module also supports the construction of a custom training library, the steps of which include: It features an open, visual editing interface, allowing therapists to upload training materials in the form of images, audio, or text. It receives custom settings from therapists, including question rules, difficulty parameters, training duration, and feedback modes, and generates training task logic. The training materials are bound and encapsulated with the training task logic to generate a personalized training library; When matching an initial rehabilitation training plan, the personalized training library is included in the pool of optional training plans, supporting direct invocation and execution.

6. The speech and cognitive impairment assessment and rehabilitation training system according to claim 1, characterized in that: The steps for generating the full-process traceability data include: Throughout the assessment and training process, the system simultaneously captures patients' touch clicks and drag-and-drop connection operation data, microphone-collected voice recording data, handwriting input trajectory data, and the answer results and reaction time for each question, forming multi-source data according to the preset collection frequency. The multi-source data is timestamped to form a raw process dataset with a single task dimension; The original process dataset is classified and archived according to four dimensions: patient unique medical record identifier, assessment / training program, execution date, and rehabilitation stage, to generate the full-process traceability data.

7. The speech and cognitive impairment assessment and rehabilitation training system according to claim 3, characterized in that: The localized Chinese assessment question bank includes a verbal assessment question bank and a cognitive assessment question bank, wherein the assessment dimensions of the cognitive assessment question bank include: The assessment includes eight neurocognitive dimensions: visuospatial construction, executive conversion, naming memory, attention to language, verbal fluency, abstract thinking, delayed recall, and spatiotemporal orientation. When performing cognitive assessment tasks, patient response data for the corresponding dimensions are collected, scores for each dimension are calculated, cognitive function sub-assessment results are generated, and the assessment results data are incorporated into the assessment results data.

8. The speech and cognitive impairment assessment and rehabilitation training system according to claim 1, characterized in that: The adaptive training module also supports multimodal interaction and adaptation to specific diseases, including the following steps: It provides independent switches and combination configuration options for visual and auditory feedback, allowing patients to select the appropriate feedback mode based on their visual and auditory impairments; For patients with unilateral neglect, the training information bar can be freely switched between the left and right sides of the interface to adapt to the patient's visual field defect. It provides customizable adjustment functions for training background color temperature and interface layout to match the needs of cognitive psychology rehabilitation environment; It has three built-in training systems for adults, the elderly and children, respectively corresponding to general speech and cognitive rehabilitation, sensory stimulation rehabilitation for the elderly and attention training for children. The corresponding training system is matched according to the patient's age and symptoms.

9. The speech and cognitive impairment assessment and rehabilitation training system according to claim 1, characterized in that: The steps for generating the comprehensive rehabilitation report include: Retrieve the corresponding patient's assessment results data, phased training performance data, and a summary of the entire process traceability data; Data from multiple modules and time periods are integrated and summarized, and the report content is organized according to the structure of patient basic information, assessment scores, training plan, efficacy trend and therapist suggestions; It supports importing custom report templates, matching and filling the integrated report content into the corresponding positions in the template, and generating the comprehensive rehabilitation report.

10. The speech and cognitive impairment assessment and rehabilitation training system according to claim 1, characterized in that: The steps for outputting the research dataset include: Using the patient's unique medical record identifier as an index, extract the patient's historical assessment and training longitudinal data; The longitudinal data is grouped and organized according to the time dimension, and the data field format and statistical caliber are standardized. A research dataset with a preset target format is generated. The dataset fields are adapted to statistical analysis software, and the dataset supports batch extraction of research data from multiple patients to form the research dataset.

Images