基于压电振动传感器的骨折愈合状态模拟实时检测系统
Real-Time Detection System for Simulating Fracture Healing Status Based on Piezoelectric Vibration Sensors
谢成花 1陈向东 1丁星 1马立泰2
作者信息
- 1. 西南交通大学信息科学与技术学院,四川 成都 610031
- 2. 四川大学华西医院骨科脊柱外科中心,四川 成都 610041
- 折叠
摘要
为解决传统影像检查无法实时检测骨折愈合状态,以及植入式检测方法手术难度大、可能引起排斥反应的问题,设计了一种基于压电振动传感器的骨折愈合状态模拟实时检测系统.该系统主要包括5个部分:骨折愈合状态模拟、振动模块、振动信号调理模块、单片机控制模块和上位机.通过模拟环境采集骨折愈合状态的振动信号,然后通过单片机控制模块传输至上位机进行处理并实时检测.同时,对比分析朴素贝叶斯、决策树和随机森林集成学习几种不同算法在检测分类效果上的差异.结果表明,使用随机森林集成学习算法进行骨折愈合状态模拟的检测分类,响应时间不超过0.2 s,准确率达98.2%,这不仅保证了系统的实时性,而且提高了系统检测分类的准确性.
Abstract
To address the issues of traditional imaging methods being unable to detect the healing status of fractures in real-time,as well as the difficulties and potential rejection reactions associated with implantable detection methods,a real-time fracture healing status simulation detection system based on a piezoelectric vibration sensor was designed.The system main includes five components:Fracture healing status simulation,vibration module,vibration signal conditioning module,microcontroller control module,and host computer.Vibration signals under different fracture healing states are collected in a simulated environment and then transmitted to the host computer for processing and real-time detection through the microcontroller control module.Additionally,a comparative analysis of several different algorithms including Naive Bayes,decision tree,and random forest ensemble learning was conducted to evaluate their effectiveness in detection and classification.The results indicate that using the random forest ensemble learning algorithm for fracture healing status simulation detection achieves a response time of no more than 0.2 s,with an accuracy of 98.2%.This not only ensures the real-time capability of the system but also enhances the accuracy of the detection and classification process.
关键词
骨折愈合状态模拟/压电振动传感器/信号处理/实时检测Key words
simulating fracture healing status/piezoelectric vibration sensors/signal processing/real-time detection引用本文复制引用
出版年
2024
NETL
NSTL
万方数据