基于FPGA的无创伤血液成分光谱数据采集系统设计

发布时间：2018-04-29 18:34

本文选题：近红外光谱 + 无创伤血液成分检测　；参考：《中国科学院研究生院(长春光学精密机械与物理研究所)》2015年硕士论文

【摘要】：血液成分检测是诊断疾病的重要手段。常规的血液成分检测采用抽血的方法,不但给病人带来痛苦和交叉感染的风险,而且不能实现连续监测。因此,人体血液成分的“无创、连续监测”无疑为人们所迫切企盼。近红外光谱在无创伤血液成分检测方面具有很大的优势,如无试剂、无污染、无创伤、可连续监测。然而,国内外学者在此领域耗费了大量的时间和精力,依然不能使近红外无创伤血液成分检测技术达到临床应用的水平。主要问题在于:血液中葡萄糖、胆固醇、甘油三酯等组分的信号非常微弱,而且血液的近红外光谱受到人体组织背景的严重干扰。为克服背景干扰,陈星旦提出了“血流容积差光谱相减法”,该方法对光谱仪器的数据采集系统提出了如下要求:(1)采集系统必须能够同时采集所有波长点处的信号;(2)采集系统每秒钟至少需采集50幅光谱图;(3)采集系统的信噪比要显著高于10000:1。依据实验室前期研究成果,采用16元阵列式探测器同时采集16个独立像元的方法,既保证所有波长点的光强信号采集自同一时刻,又能使仪器具有较高的采集速度和信噪比。这就要求仪器的数据采集系统能够同时对16个通道的血液脉搏波数据进行高速、高信噪比的采集。本文针对无创伤血液成分检测仪器中的基于FPGA的高速、多通道光谱数据采集系统进行了研究。文章主要包括光谱数据采集系统的硬件设计和FPGA逻辑设计,具体研究内容和主要结论如下:1)对数据采集系统的硬件部分进行了详细设计,用高精度A/D芯片实现16通道数据的并行采集;采用SRAM芯片实现数据储存;采用USB芯片CY7C68013实现数据通信,采用FPGA作为系统的控制核心,实现对系统各个部分的控制。2)对数据采集系统的FPGA程序进行了详细设计,包括A/D控制模块、乒乓RAM控制模块、SRAM控制模块和USB数据通信模块,并将各模块综合成完整的数据采集系统控制程序,实现了16通道人体血液脉搏波数据的高速、高信噪比采集。3)使用该系统对一个电压为1.357V的稳压源进行了测量,在19531Hz采样率下,系统信噪比可以达在40000:1。在该采样率下,对一名男性进行了脉搏波信号的采集,系统每秒采集305幅光谱图,实验证明,该系统可以采集到高信噪比的人体血液脉搏波信号。满足无创伤血液成分检测仪器对于其光谱数据采集系统的要求。4)研究了脉搏波信号滤波处理的方法,采用小波分析的方法进行脉搏波信号去噪可以获得较好的效果。通过多组对比实验表明,使用sym8小波进行两层分解,函数形式采用rigrsure,门限阈值采用硬阈值可以获得最好的滤波效果,滤波后的量化信噪比可达151.1331,均方误差为5.61E-05。
[Abstract]:The detection of blood components is an important method in the diagnosis of diseases. The method of blood sampling not only brings the patient the risk of suffering and cross infection, but also can not realize continuous monitoring. Therefore, the non-invasive, continuous monitoring of human blood composition is unquestionably expected by people. Near-infrared spectroscopy (NIR) has great advantages in the detection of non-invasive blood components, such as no reagent, no pollution, no trauma, and can be continuously monitored. However, scholars at home and abroad have spent a lot of time and energy in this field, and still can not make the near infrared non-invasive blood component detection technology to the level of clinical application. The main problem is that the signals of glucose, cholesterol and triglyceride in blood are very weak, and the near infrared spectrum of blood is seriously disturbed by the background of human body. In order to overcome the background interference, Chen Xingdan put forward the "volume difference spectral phase subtraction of blood flow". This method puts forward the following requirements for the data acquisition system of the spectrometer: 1) the acquisition system must be able to collect signals at all wavelength points at the same time.) the acquisition system needs to collect at least 50 spectral maps per second.) the signal to noise ratio of the acquisition system is significantly higher than that of 10 000: 1. According to the previous research results of the laboratory, the 16 element array detector is used to collect 16 independent pixels simultaneously, which can not only ensure that the light intensity signals of all wavelength points are collected from the same time, but also make the instrument have higher acquisition speed and signal-to-noise ratio (SNR). This requires that the data acquisition system of the instrument can simultaneously collect the blood pulse wave data of 16 channels at high speed and high signal-to-noise ratio (SNR). A high speed multichannel spectral data acquisition system based on FPGA is studied in this paper. This paper mainly includes the hardware design and FPGA logic design of the spectral data acquisition system. The specific research contents and main conclusions are as follows: (1) the hardware part of the data acquisition system is designed in detail. High precision A / D chip is used to realize 16 channel data parallel acquisition, SRAM chip is used to realize data storage, USB chip CY7C68013 is used to realize data communication, and FPGA is used as the control core of the system. The FPGA program of the data acquisition system is designed in detail, including the A / D control module, the ping-pong RAM control module and the USB data communication module. The modules are integrated into a complete control program of the data acquisition system, and the high-speed and high signal-to-noise ratio (SNR) acquisition of 16-channel human blood pulse wave data is realized. The system is used to measure a steady voltage source with a voltage of 1.357 V. under the 19531Hz sampling rate, The SNR of the system can reach 400: 1. At this sampling rate, the pulse wave signal of a male was collected. 305 spectral images per second were collected by the system. The experimental results show that the system can collect the pulse wave signal of human blood with high signal-to-noise ratio (SNR). The method of pulse wave signal filtering and processing is studied in this paper. The method of pulse wave signal de-noising by wavelet analysis method can get better results. The comparison experiments show that the best filtering effect can be obtained by using sym8 wavelet and Rigrsure. the quantization signal-to-noise ratio can reach 151.1331 and the mean square error is 5.61E-05.
【学位授予单位】：中国科学院研究生院(长春光学精密机械与物理研究所)
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：R446.1;TP274.2

【相似文献】