摘要
提出了基于生物电阻抗测量技术的血型检测新方法并研制了检测系统。该系统采用单片机结合AD9850芯片实现 DDS 技术产生激励信号,设计椭圆滤波器消除信号干扰;运用低成本放大器 AD8370设计稳幅电路,确保信号输出幅度不变;系统采用 AD8302幅相检测芯片实现阻抗信息提取;基于人工神经网络判别红细胞悬浮液凝集强度,并结合判别规则得到血型结果。测试结果表明该系统精度高,稳定性好;临床实验表明,该系统能正确识别红细胞6种不同凝集程度,能100%正确地检测出正常标本的 ABO 和 Rh 血型,可更好地满足临床需求。
A new method for blood type detection and a new automated blood typing system are developed. In the system,the combined use of the AD9850 chip and MCU are employed to realize the DDS technique and produce excitation signal. Then an elliptic filter is designed to eliminate the signal interference,and an amplitude stabilized circuit is designed to stabilize the amplitude of the signal output by using an AD8370 chip,which is a low-cost variable gain amplifier. The amplitude and phase measurement chip AD8302 is adopted to extract the impedance information. Based on neural network recognition model,a new method is designed to determine the degree of agglutination in red blood cell suspension. The use of this method combined with the discriminant rule could successfully identify the blood type. Contrastive test is carried out and the new system showed high measurement accuracy and good stability,could better distinguish the different intensity of agglutination in red blood cell,and identified ABO and Rh blood type correctly in normal samples. The self-developed new automated blood typing system is in better agreement with clinical requirements.
出处
《计量学报》
CSCD
北大核心
2014年第4期403-408,共6页
Acta Metrologica Sinica
基金
重庆市自然科学基金(CSTC2006BB3176)
重庆市卫生局科研项目(2010-2-086)
重庆大学“211工程”三期创新人才培养计划(S-09106)
关键词
计量学
生物电阻抗测量技术
神经网络
血型
红细胞悬浮液
Metrology
Bioelectrical impedance analysis technology
Neural networks
Blood type
Red blood cell suspension