Method And Experiments To Maintain Measurement Condition Of Non-invasive Sensing By Near-infrared Diffuse Reflectance

As non-invasive sensing by near-infrared(NIR)diffuse reflectance is easily to be disturbed by the interferers such as the measurement system drift,contact pressure and thermal transmission between the probe and the skin tissue,the repeatability of measurement in vivo is poor.In this thesis,from the prospective of measurement condition,the physical models to evaluate the influence of these factors on the NIR diffuse reflectance in vivo was established;and the solution to maintain the measurement condition was proposed and validated by experiments.Firstly,considering the surface of the standard diffuse reflection plate commonly used in reference measurement approximately as a series of micro face elements,the physical model of diffuse reflectance using the standard diffuse reflection plate as the reference based on Gauss distribution function was established.The influences of coarseness and positioning precision of standard plate,irradiation surface area and the distance between standard plate and the probe were quantitatively analyzed.Further,the optimization design of the reference measurement system for the detection of NIR diffuse reflectance in vivo was determined and verified by experiments.Secondly,based on the solid-liquid two-phase mechanic feature of skin tissue,the relationships between probe-tissue contact pressure and the variations in tissue thickness,water content and optical parameters were deduced.The mechanism of how the pressure affects the diffuse reflectance was analyzed by Monte Carlo simulation.Then the linear relationship between the contact pressure and thickness strain of skin tissue was verified by the in vivo experiments.What’s more,the method to estimate the contact state between probe and skin tissue was developed on the basis of 2D correlation spectroscopy and radar map,which realize the conversion between skin’s thickness strain and contact pressure.Then the in vivo experiments about the influence of contact pressure on the diffuse reflection measurement were conducted to validate the rationality of theory analysis.Moreover,the optimal pressure range in the contact measurement of NIR diffuse reflectance was determined and verified by experiments in vivo.Thirdly,the simplified formula was introduced to calculate the surface temperature of skin tissue in the state of skin-probe contact.The influences of the material and the initial temperature of the probe on the variation of skin surface temperature were analyzed quantitatively by theory and experiments in vivo.Then an optimization scheme by selecting the metal to protect the probe and adding pre-heating process was proposed to reduce the temperature variation after contact and shorten the time cost of heat transmission.For the NIR non-invasive diffuse reflectance measurement system,the most appropriate pre-heating temperature for the surface of the probe and the best measurement time window were determined and verified by experiments in vivo.Finally,control apparatus to maintain the measurement conditions was designed and built.Moreover,clinical experiments to detect the glucose concentration noninvasively by NIR contact measurement were performed,and the calibration model based on partial least square was established by relating the diffuse reflectance with the blood glucose concentration.Experiment results indicated that,the correlation coefficient between the predicted glucose and its true valuewas larger than 0.89,and the root mean square error of prediction is less than 1 mmol/L after adopting the measurement condition-controlling apparatus.These results demonstrate that the measurement condition maintaining scheme proposed in this thesis is helpful to improve the accuracy of contact measurement by non-invasive NIR diffuse reflectance.