【作者】 常凤香；

【导师】 洪文学；

【作者基本信息】 燕山大学 ， 仪器科学与技术， 2012， 博士

【摘要】 生物电阻抗测量技术作为一种无损伤检测技术，在生物医学领域中具有广泛的应用，而经穴电阻抗特性测量作为生物电阻抗测量中的一个研究方向，有着重要的理论意义和临床价值。按照中医理论，穴位可以反映人体的生理或病理特性，是经络穴位研究的主要对象。因此，本文应用生物电阻抗测量技术，以人体井穴、原穴和合穴为研究对象，研制了一套经络和穴位电阻抗测量系统，并利用该系统对人体经穴的电阻抗特性进行研究，提出了经络诊断数据可视化方法，不仅为经络诊断提供了一种新的方法，也为进一步研究人体经穴电阻抗特性与疾病之间的关系奠定了基础。首先，在对生物电阻抗测量原理深入研究的基础上，设计并构建了经络和穴位电阻抗测量系统平台。完成了系统的硬件设计和软件编写，系统支持直流测量和交流测量两种阻抗测量模式，具有较高的测量精度和较好的稳定性，而且软件的编写具有较好的扩展性。针对阻抗测量中，信号难免会受到噪声污染的问题，构造了一种新的小波阈值去噪函数，与传统的软、硬阈值相比，该函数不仅克服了硬阈值函数不连续的缺点，还克服了软阈值函数中估计小波系数与分解小波系数之间存在的恒定偏差，具有软、硬阈值不可比拟的灵活性。用该函数对阻抗测量数据处理效果理想。其次，研究了基于直流电特性的经络诊断数据可视化方法。提出用多元图表示穴位诊断数据，实现了诊断结果的可视化。实验以原穴为测量穴位并进行了病例分析，结果表明用该方法诊断的结果与患者病症相吻合。提出将测量的电阻抗通过彩色图像的方式直观的显示出来，不同的颜色代表不用的电阻率，反映穴位不同的信息，进而推断人体的健康状况，采用主成分分析提取经络特征，并用BP神经网络实现经络特征参数的辨识。实验以亚健康人群为研究对象，结果表明，该方法可以对样本进行有效地识别。最后，研究了基于阻抗频谱特性的经络诊断数据可视化方法。以最小二乘法提取Cole-Cole模型中的各个参数，以人体曲池穴及其对照点为测量点，对经络阻抗频谱特性进行了实验研究。并在此基础上，提出了一种基于阻抗频谱特性的经络诊断数据可视化方法，实现了多穴位、多频率测量的经络诊断数据的可视化。按照不同频率对生成的图像切割，还可以研究单频率点的阻抗特性。实验结果表明，用该方法不仅实现了高维阻抗数据的可视化，还可以对健康样本和疾病样本进行有效区分。更多还原

【Abstract】 As a non-destructive testing technology, bioelectrical impedance measurement hasbeen widely used in biomedical field, while measuring electrical impedance of meridianand acupuncture points, as a research direction of bioelectrical impedance measurements,has important theoretical significance and clinical value. According to Chinese medicinetheory, acupuncture points can reflect the physiological or pathological features of thehuman body, which make them become the mainly study content in the filed of themeridian and acupuncture points research. Therefore, application of bioelectricalimpedance measurement technique, an impedance measuring system was developed inthis paper. The Well, Source and Sea acupuncture points were chosen as detection points,the electrical impedance characteristics of meridian and acupuncture points wasinvestigated with this system and the meridian diagnostic data visualization method wasproposed, which not only offer a new method for meridian diagnosis, but also load thefoundation for the further study the relationships between the electrical impedancecharacteristics of meridian and acupuncture points and the diseases of the human body.Firstly, an impedance measuring system was designed and constructed based on thebioelcetrical impedance measuring principle, and the hardware and software designs werealso completed. The system supported DC and AC measuring modes with high accuracyand good stability, and the software design had great expandability. For the impedancemeasurement, the signal spectrum and noise spectrum having great overlap, so a newwavelet threshold denoising function was structured. Compared with soft- andhard-thresholding function, the new thresholding function overcome the discontinuity ofthe hard-thresholding method and reduced the fixed deviation between the estimatedwavelet coefficients and the decomposed wavelet coefficients of the soft-thresholdingmethod, it was more elastic than soft- and hard-thresholding function. The newthresholding function was used to impedance data processing, and the desired results wereobtained.Secondly, research on the meridian diagnostic data visualization method based on DCcharacteristics of the meridian and acupuncture points. The diganostic data of the acupuncture points represented by the multidimensional chart was proposed, and thediagnostic results visualization was achieved. The Source points were chosen asmeasuring point in the experiment, and a case was analyzed, the results showed that thediagnostic results were consistent with the symptoms of the patient. The impedance datadisplayed through the color image was proposed, different colors represented differentresistivity to reflect different information of acupuncture points, and then the health stateof human body was infered. Principal component analysis was used to extract the meridiancharacteristics and BP neural network was used to achieve the identification of themeridian characteristic parameters. The sub-health people were chosen as research objectsin the experiment, and the results showed that the samples could be identified effectivelywith this method.Lastly, research on the meridian diagnostic data visualization method based onimpedance spectrum characteristics. The least squares method was used to extract theparameters of the Cole-Cole model, Quchi and its control point were chosen as measuringpoints, and the impedance spectral characteristics of the meridian and acupuncture pointswas studied. Based on this, a method of meridian diagnostic data visualization wasproposed, which achieved the visualization of the multi-points, multi-frequencymeasurements of the impedance data. The single frequency impedance characteristicscould be studied through image cutting. The experimental results showed that the methodnot only achieved the high dimensional impedance data visualization, but alsodistinguished effectively between healthy samples and disease samples.更多还原