用于漫射光检测的光子计数锁相检测系统与实验研究

发布时间：2018-05-22 07:23

本文选题：多波长稳态漫射光检测 + 光子计数技术　；参考：《天津大学》2012年硕士论文

【摘要】：稳态扩散光学成像(Continuous Wave Diffuse Optical Imaging, CW DOI)利用近红外光在组织体内相对较深的穿透深度实现器官级的人体信息检测,其优点包括无创、无电离、时间分辨率高以及适合临床连续监测等。CW DOI通常需要测量组织体在不同光波长从多个源点激励下表面其它各点的漫反射光流分布,由此带来光源波长与源点位置编码的问题。当前大多数CW DOI系统采用时分复用光源编码方式,该方式测量时间长难以满足某些快速变化生理信号实时监测的要求。为此,本文提出了光源频分复用编码与探测器输出信号数字锁相检测解复用的实现方案,减少了测量时间,提高了测量结果的时间分辨率。围绕上述方案所搭建的稳态双通道漫射光检测系统包括光源、波分复用器、光子计数探测器、FPGA数字锁相检测电路、通信与数模转换电路以及上位机软件部分。实验发现,该系统存在测量暗噪声较大,测量结果信噪比不够好,入射光强与源探距离均不变的情况下测量结果间或出现粗大误差现象以及FPGA检测系统与PC机上位机之间数据传输速率不满足要求等问题。针对以上问题,本文分别进行了原因分析、改进方案设计以及改进效果的实验验证。针对暗噪声高与信噪比不够好的问题,研究了实验条件对测量结果的影响,采用了纤芯更细的光纤;针对测量结果中间或出现的粗大误差现象进行了前端探测器输出信号的阻抗匹配以及FPGA输入异步信号进行了同步化处理;针对数据传输速率不满足要求的问题,采用向量模式的CORDIC算法在FPGA内部实现了均方根运算模块,以此减轻上位机工作负担且减少了实际需传输的数据量,提高了有效数据传输速率。最后针对以上改进方案进行了实验验证。结果表明,改进后的系统在相同源探距离下的测量值变化规律与实际入射光强变化规律基本一致,二者最大相对误差为7.32%;在更大源探距离范围内测量结果的信噪比达到了35dB以上;在16-30mm源探距离范围内,双通道测量值与解析解结果变化规律基本一致,相对误差保持在17%以内。改进后的系统暗噪声测量值明显减小,在测量准确度与稳定性均有较大幅度提高。
[Abstract]:Steady-state diffusion optical imaging continuous Wave Diffuse Optical Imaging, CW DOI) uses near-infrared light to detect organ-level human information at a relatively deep penetration depth in the tissue. Its advantages include non-invasive and non-ionizing. High temporal resolution and suitable for clinical continuous monitoring. CW DOI usually needs to measure the diffuse reflected light flow distribution of tissue at different light wave lengths from multiple source points, which leads to the problem of source wavelength and source position coding. At present, most CW DOI systems adopt time division multiplexing (TDM) light source coding method, which is difficult to meet the requirements of real-time monitoring of some rapidly changing physiological signals. Therefore, this paper proposes a scheme of demultiplexing for the frequency division multiplexing coding of light source and digital phase-locked detection of detector output signal, which reduces the measurement time and improves the time resolution of the measurement results. The steady-state dual-channel diffuse light detection system built around the above scheme includes light source, wavelength demultiplexer, photon counting detector FPGA digital phase-locked detection circuit, communication and digital-to-analog conversion circuit and upper computer software. The experimental results show that the system has large dark noise, and the signal to noise ratio is not good enough. Under the condition that the incident light intensity and the source detection distance are both invariable, there may be a gross error between the measurement results and the data transmission rate between the FPGA detection system and the PC computer. In order to solve the above problems, this paper analyzes the reasons, designs the improved scheme and verifies the effect of the improvement. Aiming at the problem of high dark noise and insufficient signal-to-noise ratio (SNR), the influence of experimental conditions on the measurement results is studied. The fiber with finer core is used. The impedance matching of the output signal of the front-end detector and the synchronous processing of the FPGA input asynchronous signal are carried out in view of the gross error phenomenon in the middle or the middle of the measurement results, and the problem that the data transmission rate does not meet the requirements is discussed. The CORDIC algorithm of vector pattern is used to implement the root mean square (RMS) operation module in FPGA, so as to reduce the workload of upper computer, reduce the amount of data needed to be transmitted, and improve the effective data transmission rate. Finally, the improved scheme is verified by experiments. The results show that the measured values of the improved system under the same source range are basically consistent with those of the actual incident light intensity. The maximum relative error is 7.322.The signal-to-noise ratio (SNR) of the measured results is above 35dB in the larger range of source range, and in the range of 16-30mm source detection, the law of variation between the measured values of two channels and the analytical solution is basically the same, and the relative error is kept within 17%. The measured value of dark noise of the improved system is obviously reduced, and the accuracy and stability of the system are greatly improved.
【学位授予单位】：天津大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：R318.51

【参考文献】