OBJECTIVE To explore optimization pathways for the drug traceability code management model in outpatient pharmacy workflows， providing practical evidence for enhancing the efficiency of pharmaceutical service. METHODS Taking the outpatient pharmacy of the First Affiliated Hospital of Sun Yat-sen University as the research subject， a comprehensive drug traceability system was established through three key interventions： upgrading the information system architecture [including integration of the hospital information system （HIS） with the traceability platform]， workflow optimization （reorganizing the inventory-dispensing-verification tripartite process）， and designing a dual-mode traceability data collection mechanism （primary data capture at dispensing stations and supplementary capture at verification stations）. Operational efficiency differences before and after implementation were analyzed using the medical insurance data and service timeliness metrics in September 2024. RESULTS After the implementation of drug traceability code management， in terms of data collection: Mode Ⅰ （verification-stage capture） uploaded 26 144 records， while Mode Ⅲ （inventory-as-sales capture） uploaded 443 061 records， totaling 469 205 entries； in terms of time efficiency： average drug dispensing time increased from 28.74 s to 43.37 s （enhanced by 51%）. Through dynamic staffing adjustments， patient wait time only extended from 8.04 min to 8.67 min （enhanced by 8%）. CONCLUSIONS Drug traceability code management can be effectively implemented via a “system reconstruction-process reengineering-human-machine collaboration” trinity strategy， leveraging informatization （e.g.， dual-mode data capture） to offset manual operation delays， which validates the feasibility of balancing national traceability demands with service efficiency in outpatient pharmacies.

This paper reviews the published articles of recent years on the application of deep learning methods in automatic localization of acupoint, and summarizes it from 3 key links, i.e. the dataset construction, the neural network model design, and the accuracy evaluation of acupoint localization. The significant progress has been obtained in the field of deep learning for acupoint localization, but the scale of acupoint detection needs to be expanded and the precision, the generalization ability, and the real-time performance of the model be advanced. The future research should focus on the support of standardized datasets, and the integration of 3D modeling and multimodal data fusion, so as to increase the accuracy and strengthen the personalization of acupoint localization.
Deep Learning ; Acupuncture Points ; Humans ; Neural Networks, Computer

Objective This study aimed to develop an experimental animal platform for evaluating the feasibility and safety of intelligent acupuncture robots and to lay the foundation for further research.Methods Six 2-month-old Guangxi Bama miniature pigs were used as experimental subjects for acupuncture verification after anesthesia.First,manual acupuncture verification was carried out.Six acupoints were selected for each experimental animal and the needles were left for 20 min after the lifting,inserting,and twisting manipulation.Before and after controls were included.The experiment was carried out for 28 days,and each experiment was conducted once every 2 days for a total of 10 times.After verification of manual acupuncture,a point 10 mm from each of the six selected acupoints was selected,with a total of 12 points,and acupuncture operations were carried out on the experimental animals using the intelligent acupuncture module of the acupuncture robot at different frequencies and angles,to further verify the stability and feasibility of the animal platform.Results Routine safety-related blood indicators and blood biochemistry indicators after the procedure were normal and stable compared with those before the procedure.The average heart rate of the animals was 124 beats/min,the average blood pressure was 87/36 mmHg,and the average body temperature of was 36℃at a room temperature of 25℃,with no significant change in body temperature during and after the experiment.On the basis of this experimental platform,acupuncture manipulation using the intelligent acupuncture module of the acupuncture robot was completed successfully,with no abnormalities related to acupuncture such as bending,breaking,or stagnation of needles during the experimental process,and the experimental animals showed no obvious abnormalities.Conclusions This study established a stable experimental animal platform for evaluating the feasibility and safety of acupuncture carried out by intelligent acupuncture robots,based on the existing experimental method of miniature pigs.These result lay a foundation for further research related to the use of intelligent acupuncture robots.

OBJECTIVE: To evaluate the performance of a newly developed intelligent acupuncture robot that simulates human manual acupuncture techniques through an animal experiment using miniature pigs. METHODS: Two 3-month-old Bama miniature pigs were selected. One pig was used for the manual needling techniques of the practitioner, and the other pig was used for the intelligent acupuncture robot. The acupoints selected were "Qiangfeng" "Bojian" "Xiaokua" "Huiyang" (BL 35) and "Baihui" (GV 20), with a straight insertion depth of 25-35 mm. The manipulation techniques, including lifting-thrusting and twisting at a frequency of 90 times/min, were applied for 1 min, with 30 s each for lifting-thrusting (10 mm amplitude) and twisting (180° angle). The practitioner's needling techniques were captured using an optical motion capture system, and the collected data were input into the intelligent acupuncture robot for replication on the second miniature pig. A complete needling session of five acupoints was considered one round, and three rounds were performed per session, every other day, for a total of five sessions, completing 15 rounds. The Fréchet distance method was used to compare the robot's replicated needling data with the practitioner's data, and curve fitting analysis was conducted to evaluate the robot's performance and precision. RESULTS: The intelligent acupuncture robot achieved a total completion rate of 96.00% for 75 needling operations, with a non-completion rate of 4.00%. The robot's replication of the practitioner's needling techniques showed a good fit, with many characteristic points overlapping. The average Fréchet distance was 18.67. The Fréchet distance for the lifting-thrusting and twisting techniques at Baihui (GV 20) and "Huiyang" (BL 35) acupoints was smaller than that at "Xiaokua" and "Qiangfeng" points (P<0.01). The accuracy of the lifting-thrusting technique at "Xiaokua" and "Qiangfeng" acupoints was lower than the twisting technique (P<0.01). CONCLUSION The intelligent acupuncture robot is able to replicate the pre-recorded manual needling parameters with a high degree of accuracy.
Animals ; Swine ; Swine, Miniature ; Acupuncture Therapy/methods* ; Robotics/instrumentation* ; Humans ; Acupuncture Points ; Female ; Male