A traditional Chinese medicine shop skill integrated training method and system based on multi-modal perception and closed-loop feedback
Patent Information
- Application Number
- CN202610659512.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-05-13
- Publication Date
- 2026-10-27
AI Technical Summary
[0009]本发明旨在解决现有培训中周期长、实操弱、考核模糊的问题，提供一种标准化、高效化、零差错的技能一体化培训方法及系统
[0036]
| 处方审核准确率 | 85% | 96.8% | 提升11.8个百分点 |
| 调剂效率 | 15分钟/处方 | 6.8分钟/处方 | 提速约55% |
| 饮片鉴别错误率 | 12% | 0.7% | 下降幅度达94% |
| 蒙眼盲辨正确率 | 约32% | ≥95% | 提升约3倍 |
| 远期用药安全 | — | 培训后3个月零不良事件 | 实现零差错安全目标 |
Smart Images
Figure CN122989360A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of traditional Chinese medicine pharmaceutical services and vocational skills training technology. Specifically, it relates to an integrated training method and system for new pharmacists in traditional Chinese medicine pharmacies based on multimodal perception and closed-loop feedback. It covers a full-chain skills training system including sensory identification of medicinal slices, prescription review, dispensing accuracy training, and process innovation and optimization. Background Technology
[0002] The following are some of the most prominent problems in the current training of new pharmacists in traditional Chinese medicine pharmacies:
[0003] First, the training period is long and the learning curve is slow. The traditional master-apprentice model relies on personal experience and takes an average of 3 to 6 months, lacking standardized and reproducible processes.
[0004] Second, the ability to identify medicinal materials is weak. Training on easily confused medicinal materials (such as angelica and angelica pubescens) relies solely on visual observation, lacking a comprehensive training system that integrates touch and smell, resulting in a high rate of dispensing errors.
[0005] Third, prescription review lacks practical experience. The identification of risk points such as the "Eighteen Incompatibilities" and "Nineteen Antagonisms" is mostly theoretical in teaching, lacking practical training with a large database of real prescriptions.
[0006] Fourth, the methods for improving precision training are limited. Techniques such as weighing with mechanical scales and packaging mulberry bark paper lack quantitative standards, posing a risk of extinction for the inheritance of traditional crafts.
[0007] Fifth, the evaluation system is subjective and vague. The assessment is mainly based on written tests, lacking quantitative indicators for the entire process and long-term effect tracking. Summary of the Invention
[0008] I. Purpose of the Invention
[0009] This invention aims to solve the problems of long training cycles, weak practical skills, and vague assessments in existing training programs, and to provide a standardized, efficient, and error-free integrated skills training method and system.
[0010] II. Technical Solution
[0011] 1. A method for integrated training of skills in traditional Chinese medicine pharmacies based on multimodal perception and closed-loop feedback, comprising the following steps:
[0012] Step S1: Construct a three-tiered, progressive teaching system
[0013] The training process is divided into three progressive stages: basic cognitive level, skills practice level, and innovation optimization level. Each stage has clearly defined teaching objectives, training content, and achievement standards. See the table below for details:
[0014] Basic cognitive level Week 1 Master the basic identification skills of medicinal slices and the rules of compatibility and incompatibilities. Studying the compatibility charts of medicinal herbs, comparing and observing genuine and counterfeit medicinal herbs, and marking the compatibility of toxic medicinal herbs with red lines. Skills Practice Level Weeks 2-3 Improve prescription review and dispensing skills Practical training in historical prescription review, high-precision weighing with mechanical scales, and standardized operation of traditional processes. Innovation Optimization Layer Week 4 Promote process optimization and innovation transformation Process optimization workshop: proposal discussion, feasibility assessment of innovative solutions, and implementation.
[0015] Step S2: Construct a multimodal perception training module
[0016] This method comprehensively utilizes visual, tactile, and olfactory sensory channels, enhancing integrated sensory discrimination capabilities through independent training and fusion recognition. It includes the following sub-steps:
[0017] (1) Visual perception training: Using the compatibility indicative book and the genuine and fake comparison of medicinal slices, train students to identify texture, cross-sectional features and color.
[0018] (2) Tactile perception training: The blindfolded identification training method is adopted to shield visual input and enhance students' tactile recognition of texture, hardness and fiber.
[0019] (3) Olfactory perception training: The blindfolded identification method is also used to identify the variety and quality of medicinal slices through odor characteristics.
[0020] (4) Multimodal fusion identification: integrate visual, tactile and olfactory information to make accurate identification of easily confused medicinal slices.
[0021] Step S3: Construct a standardized prescription review training module
[0022] (1) Construction of a knowledge graph of incompatibilities: Establish a structured knowledge graph that includes rules such as the eighteen incompatibilities, the nineteen antagonisms, and the red lines for compatibility of toxic medicinal slices.
[0023] (2) Construction of historical prescription database: collect and organize no less than 100 typical prescriptions, covering various risk scenarios.
[0024] (3) Red and blue label dual-color review training: Trainees review historical prescriptions, use red pen to circle risk points (such as incompatibilities), and use blue pen to mark legal alternatives (such as safflower → salvia miltiorrhiza).
[0025] Step S4: Construct the accuracy adjustment training module
[0026] (1) High-precision weighing training with mechanical scales: Use a scale with an accuracy of not less than 0.1g and set a quantitative target.
[0027] (2) Standardized training in traditional crafts: including training in the standardized operation procedures of the "three-fold and one-press" packaging method of mulberry bark paper and the pottery jar boiling process.
[0028] Step S5: Construct a closed-loop feedback evaluation module
[0029] (1) Three-dimensional quantitative assessment: The three dimensions of prescription review accuracy, dispensing precision and efficiency of medicinal slices, and feasibility of innovative proposals are evaluated by weight.
[0030] (2) Personalized learning path recommendation: Generate a student's ability profile based on the assessment results and push targeted reinforcement training.
[0031] (3) Long-term tracking of training effectiveness: Record medication safety event data within 3 months after training to form a long-term verification closed loop.
[0032] 2. An integrated training system for traditional Chinese medicine pharmacy skills based on multimodal perception and closed-loop feedback, comprising the following modules:
[0033] (1) Layered teaching management module; (2) Multimodal perception training module; (3) Prescription standard review training module; (4) Dispensing accuracy training module; (5) Closed-loop feedback evaluation module; (6) Evaluation report.
[0034] III. Beneficial Effects
[0035] Compared with the prior art, the present invention has the following significant technical effects:
[0036] Prescription review accuracy 85% 96.8% An increase of 11.8 percentage points Adjustment efficiency 15 minutes / prescription 6.8 minutes / prescription Speed increase of approximately 55% Error rate in identifying medicinal slices 12% 0.7% The decline reached 94%. Blindfolded identification accuracy Approximately 32% ≥95% Increased by approximately 3 times Long-term medication safety — Zero adverse events in 3 months after training Achieving zero-error safety goals Attached Figure Description
[0037] Figure 1 : Overall flowchart of the training method of this invention;
[0038] Figure 2 Diagram of a three-tiered, progressive mentoring system;
[0039] Figure 3 Schematic diagram of the multimodal perception training module;
[0040] Figure 4 Schematic diagram of the blind identification training device;
[0041] Figure 5 Three-dimensional quantitative assessment weight distribution chart. Detailed Implementation
[0042] The following detailed description is based on specific embodiments.
[0043] Example 1: Implementation of Basic Cognitive Level Training (Week 1)
[0044] Ten newly hired pharmacists were selected from the herbal medicine department of a top-tier hospital for training.
[0045] 1. Multimodal Perception Training – Visual Identification of Medicinal Herbs: Conduct comparative training on genuine and counterfeit herbs in the specimen cabinet area (e.g., genuine Astragalus membranaceus has a radial "chrysanthemum heart" texture on the fracture surface, while counterfeit herbs are hard and lack texture). Select 15 groups of easily confused medicinal herbs (Angelica sinensis vs. Angelica pubescens, etc.) for comparative identification.
[0046] 2. Blindfold Identification Challenge for Medicinal Herbs: Figure 4This is a schematic diagram of the blind identification training device. The training box 1 is opaque and made of wood, with external dimensions of 50cm (length), 40cm (width), and 30cm (height). Two circular hand-holding openings 2 are located on the front of box 1, each with a diameter of 15cm. A black velvet curtain 3, composed of multiple overlapping layers of velvet, is sewn inside the openings 2 to prevent external light from entering the box. Box 1 includes a removable tray 4 for holding medicinal herbs, measuring 40cm (length) and 30cm (width). The tray is divided into multiple compartments, each holding one sample of the medicinal herb to be identified. The outer rim of the circular hand-holding openings 2 is coated with a herb-preserving odor coating 6 (made of a mixture of food-grade activated carbon and diatomaceous earth) to ensure the continuous retention of the herb's odor, making the training environment more realistic.
[0047] During training, the instructor places the sample 5 of the medicinal slices to be identified into the compartment of tray 4, and the trainee inserts both hands into the box 1 through the hand-access opening 2. Due to the blackout curtain 3, the trainee cannot see the appearance of the medicinal slices inside the box and can only rely on touch to perceive the texture (hard / soft), cross-sectional characteristics (powdery / fibrous), and shape (flake / block / filamentous) of the slices, as well as smell to perceive the aroma characteristics (strong / light, fragrant / unique, sweet / bitter, etc.), to comprehensively judge the variety of the medicinal slices. After training, the trainee can pull out tray 4 to check the correct answer, achieving immediate feedback. After training, the blindfolded identification accuracy rate increased from approximately 32% to over 95%.
[0048] Example 2: Implementation of Skills Practice Level Training (Weeks 2-3)
[0049] 1. Practical Training in Prescription Red and Blue Label Review: A historical database of 120 typical prescriptions was built (covering 25 prescriptions related to the Eighteen Incompatibilities, 25 to the Nineteen Antagonistic Prescriptions, and 20 prescriptions involving excessive toxicity). Trainees used red and blue labels for review, with instructors providing feedback on each prescription. A total of 20 hours of training was completed.
[0050] 2. Dispensing Accuracy and Efficiency Training: Using a 0.1g precision balance, complete 10 sets of prescription dispensing exercises. The training goal is to weigh 5 herbs in a single prescription within ≤7 minutes with an error of ≤0.3g. After training, the dispensing efficiency improved from 15 minutes / prescription to 6.8 minutes / prescription.
[0051] Example 3: Implementation of Innovation Optimization Layer Training (Week 4)
[0052] 1. Process optimization workshops: Weekly proposal discussions and feasibility assessments.
[0053] 2. Inheritance of Traditional Craftsmanship: The standard "three folds and one press" movement is broken down and a demonstration video is created. Trainees retain 100% of the required skills.
[0054] Example 4: Implementation of Closed-Loop Feedback Evaluation
[0055] 1. Three-dimensional quantitative assessment: A weighted assessment method is used, with a passing score of ≥90 points. Specific weights are as follows:
[0056] Prescription review 50 prescriptions will be randomly selected for review within a specified time. 35% Accuracy rate no less than 90% Herbal medicine preparation Weigh 3 sets of prescriptions (error ≤ 0.3g / ingredient) 40% The weighing pass rate shall not be less than 90%. Innovation proposal Feasibility assessment of process optimization recommendations 25% Overall score no less than 80 points
[0057] 2. Long-term follow-up: Within 3 months after the training, the incidence of adverse drug events among the 10 trainees was 0, and the refill error rate was 0.
Claims
1. A method for integrated training of traditional Chinese medicine pharmacy skills based on multimodal perception and closed-loop feedback, characterized in that, Includes the following steps: Step S1: Construct a three-level progressive tiered teaching system, dividing the training process into a basic cognitive layer, a skills practice layer, and an innovation and optimization layer; Step S2: Construct a multimodal perception training module that comprehensively utilizes the three perception channels of vision, touch, and smell, including the blindfolded identification training method, which shields visual input in an opaque training box to enhance tactile and odor recognition; Step S3: Construct a prescription review training module, establish a knowledge graph of drug incompatibilities and a historical prescription database, and use the red and blue label dual-color review method for training; Step S4: Construct an adjustment accuracy training module, set quantitative accuracy and efficiency targets, and conduct training on mechanical scale weighing and traditional processes; Step S5: Construct a closed-loop feedback evaluation module to conduct three-dimensional weighted assessment and long-term effect tracking.
2. The method according to claim 1, characterized in that, In step S1: (1) The basic cognitive level includes learning the compatibility and incompatibility chart and comparing and observing genuine and counterfeit medicinal slices; (2) The skills practice layer includes practical experience in reviewing historical prescriptions and high-precision weighing training with mechanical scales; (3) The innovation and optimization layer includes process optimization proposals and feasibility assessments.
3. The method according to claim 1, characterized in that, The multimodal perception training module in step S2 includes: (1) Visual perception training; (2) Blindfolded tactile perception training; (3) Blindfolded olfactory perception training; (4) Multimodal fusion identification.
4. The method according to claim 1 or 3, characterized in that, The blindfolded identification training is conducted in an opaque training box, which is a box with a hand entry opening on the side wall and a replaceable herbal medicine tray inside.
5. The method according to claim 1, characterized in that, The red and blue labeling dual-color review method in step S3 is as follows: use a red pen to circle the risk points and use a blue pen to mark the legal alternatives.
6. The method according to claim 1, characterized in that, The historical prescription database in step S3 contains no fewer than 100 typical prescriptions, covering risk types such as the Eighteen Incompatibilities, the Nineteen Antagonisms, and toxic overdose.
7. The method according to claim 1, characterized in that, The quantitative accuracy target in step S4 is: the weighing time for a single prescription of 5 herbs should not exceed 7 minutes, and the weighing error for each herb should not exceed 0.3g.
8. The method according to claim 1, characterized in that, The standardized training of traditional processes in step S4 includes standardized operation training for the "three-fold and one-press" packaging method of mulberry bark paper.
9. The method according to claim 1, characterized in that, The weighted assessment in step S5 is as follows: prescription review 35%, herbal medicine dispensing 40%, and innovation proposal 25%. A comprehensive score of not less than 90 points is considered qualified.
10. A system for implementing the training method according to any one of claims 1-9, characterized in that, include: The system includes a tiered mentoring management module, a multimodal perception training module (including a blind identification training device), a prescription review training module, a dispensing accuracy training module, a closed-loop feedback evaluation module, and a central data processing unit.