Design and implementation of OCR model-based data collection platform for medical treatment equipment
10.19745/j.1003-8868.2024167
- VernacularTitle:基于OCR模型的医疗救治装备数据采集平台设计与实现
- Author:
Ke-Yu FANG
1
;
Xin ZHANG
;
Jun-Jun WANG
;
Xiao-Li QIN
;
Ping CHEN
Author Information
1. 军事科学院系统工程研究院,天津 300161
- Keywords:
optical character recognition;
emergency medical rescue;
medical treatment equipment;
data collection
- From:
Chinese Medical Equipment Journal
2024;45(9):14-20
- CountryChina
- Language:Chinese
-
Abstract:
Objective To design a data collection platform for medical treatment equipment based on an optical character recognition(OCR)model to facilitate auto medical data acquisition during emergency disaster relief.Methods A data collection platform for medical treatment equipment was constructed with the"perception—network—platfrom"architecture of medical Internet of Things.Firstly,Raspberry Pi 4B was adopted as the edge node,and the hardware environment was established with the video capture card,camera and tablet computer.Secondly,the OCR model was optimized based on convolutional recurrent neural network(CRNN),and video stream processing and data extraction at the the medical terminal were realized through hardware and software collaboration.Finally,FineBI tool was used to implement interactive interface design and database link.Results The platform developed was proved by experiments to have its hardware environment gifted with high reliability and stability and the optimized OCR model with enhanced text recognition accuracy,which contributed to rapid and automatic data acquisition for medical equipment.Conclusion The platform developed provides comprehensive and accurate medical treatment equipment data support for medical personnel,which is conducive to enhancing the efficiency of medical treatment.[Chinese Medical Equipment Journal,2024,45(9):14-20]
