眼部信号检测分析及应用

发布时间：2018-06-16 01:27

本文选题：眼部 + 信号　；参考：《华中科技大学》2014年博士论文

【摘要】：从认知层面发掘目标信息,利用现有硬件系统进行信号采集,通过软件系统及特定数学算法对所拾取信号进行处理分析,并在干扰噪声中准确提取目标信号的过程,称之为信号检测分析。1954年该理论由Tanner WP首次提出。眼球作为中枢神经系统的延伸部分,其在稳定的眼内压支撑下不断将所接收的光学信号转变为电信号,为生物体提供丰富地视觉刺激。如何以现有理论认知为支撑,利用硬件及软件系统准确采集,处理及分析有用的眼部信号,是眼部信号检测分析理论实际应用中的核心及难点。本研究主要分为四大部分：1.在现有临床检测方法对昼夜眼压波动信号拾取过程具有局限性的理论支撑下,第一部分主要通过医工交叉,利用机械测量原理,建立采集、提取及分析眼内压波动信号的新方法；2.在基于眼部血流量异常直接参与视神经疾病进展过程的理论支撑下,第二,三,四部分主要利用现有无创检测技术及相应仪器(相干光断层扫描),在选取最佳测量方法后,提取正常人,屈光不正及青光眼患者眼部视网膜及脉络膜信号进行分析,处理。观察脉络膜组织与视网膜退行性变的关系及有氧运动前后正常人与原发性开角型青光眼患者脉络膜信号的反应差异性。一、基于磁斥力连续性眼压测量原理与方法研究目的：利用“永磁铁磁斥力测量”原理提取眼内压信号,有望达到24h眼内压监测的最终目的。原理：永磁铁间距渐进性缩小,磁斥力逐渐增加,当磁斥力(使力磁环与受力磁环间的排斥力)增大与眼内压平衡时,继续缩短磁间距,角膜将在外力作用下发生形变。此时在示波器所获取的电压-位移曲线中可观察到小幅波动。波动处电压输出信号(U)与眼内压(P)相关。通过建立U、P对应关系的数学模型,从压力曲线信号中提取U信号,间接可得眼压P信号大小。方法：以35mmHg作为眼压信号检测上限,利用“有限元仿真方法”评估导致角膜发生形变所需永磁铁磁斥力大小及其位移范围。建立模拟眼球灌注模型,构建机械测量装置,提取不同压力下传感器电压输出信号。结果：1.N40H稀土永磁铁材料能够达到本研究需求,产生足够磁斥力压迫角膜发生形变。2.不同压力下电压-位移曲线波动点处传感器输出电压信号不同。伴随模拟眼球内压力逐渐升高(10mmHg,30mmHg,50mmHg),压力传感器输出电压信号相应升高,二者相关性良好。结论：利用此种方法通过拾取电容压力感受器电压输出信号间接反映眼内压波动,整体测量过程无创,快速,有望达到无创24h眼压监测的最终目的。二、经眼睑眼内压检测系统的研究目的：利用“生物组织弹性力学差异原理”提取眼内压信号,有望达到使用者自测眼压的最终目的。原理：“经眼睑眼内压检测系统”基于生物组织的弹性力学差异。利用眼球不同组织间存在的“力-形变”差异获取眼压信号。方法：本研究采用机械力学“压缩测量方法”对眼睑及眼球壁(角膜组织)进行弹性压缩,通过电容压力传感器获取压力反射信号。联合位移传感器(可变电阻器)由示波器采集并绘制出电压-位移曲线,曲线“拐点”处代表不同组织交界面,拐点后曲线斜率与眼内压相关。计算拐点后曲线斜率,观察曲线斜率与眼内压相关性,完成该系统的初步研究工作。结果：1.本研究测量眼内压范围13-33mmHg,可以达到大多数原发性开角型青光眼患者眼压信号幅度。2.在不同灌注压力下,重复测量4次,实验结果示电压-位移曲线重复性良好。3.伴随眼内压逐渐升高,拐点后电压-位移曲线斜率相应增大,线性拟合后可以通过电压-位移曲线斜率K间接获得眼内压幅值。结论：经眼睑眼内压测量系统原理可行,结果重复性佳,能够实时反映眼内压变化,可以达到使用者自测眼压的最终目的。第二部分：不同方法提取脉络膜厚度信息的差异性研究目的：招募正常受试者为研究对象,运用相干光断层扫描技术提取人眼脉络膜组织信号,分别选用海德堡测量软件及ImageJ测量软件对脉络膜厚度进行测量分析,进行不同测量方法可重复性及一致性的相关统计学分析,选取出最优测量方法为后续研究奠定基础。方法：36名正常受试者纳入此项研究。利用相干光断层扫描技术采集中心凹及旁中心凹直径6mm范围内脉络膜组织图像。分别由Heidelberg eye explore software(版本5.3.3.0,海德堡,德国)及Image J software (version1.42,美国)进行脉络膜厚度测量。运用Bland-Altmann统计学方法对不同测量方法测量结果一致性及可重复性进行分析比较。结果：1. Image J软件测量重复系数(39.9186)高于海德堡测量软件(27.3525)；2.两种方法测量结果一致性分析显示,95%可信区间为-18.437～63.949um；3.95%可信区间精确性分析显示：其上限范围16.102～111.796um；下限范围-66.29～21.41umm。结论：Image J测量软件重复性及一致性较Heidelberg测量软件高,应该成为研究者日后脉络膜厚度测量的可选用软件之一。第三部分：近视患者脉络膜厚度及感光细胞层厚度变化的研究目的：通过招募正常人及近视患者作为研究人群,利用相干光断层扫描技术提取人眼脉络膜及视网膜组织信号,有望寻找脉络膜与视网膜组织退行性变间的相互关系。方法：64例受试者依据屈光状态分为三组：组Ⅰ：正视组(+1.0D～-1.OD)；组Ⅱ：轻中度近视组(-1.0D～-6.0D)；组Ⅲ：高度近视组(-6.0D)。运用频域相干光断层扫描技术(spectral domain optic coherence tomography, SD-OCT)提取脉络膜及视网膜组织图像信号。利用Image J软件提取视网膜厚度(Retinal thickness, RT),视网膜神经纤维层厚度(retinal nerve fiber layer thickness, RNFLT),视网膜神经节细胞层厚度(ganglion layer thickness, GLT),视网膜感光细胞层厚度(photoreceptor layer thickness, PRLT)及脉络膜厚度(choroidal thickness, ChT)信息。结果：与正视者相比,高度近视患者感光细胞层厚度,视网膜厚度及脉络膜厚度均显著降低(P0.05)；单因素及多因素线性回归分析方法均提示感光细胞层厚度变化与脉络膜厚度变化紧密相关。结论：高度近视患者不仅脉络膜厚度较正常人降低,感光细胞层厚度同样降低。二者构成神经血管单元。脉络膜厚度变化会导致感光细胞层厚度变化,反之亦然。目的：本研究将原发性开角型青光眼(primary open angle glaucoma, POAG)患者作为主要研究人群,利用相干光断层扫描技术(optic coherence tomography,OCT)提取受试者脉络膜组织图像信号,利用非接触式眼压计提取眼内压信号,探究有氧运动前及有氧运动后正常受试者与POAG患者眼部脉络膜厚度及眼内压力变化情况。方法：34例受试者分为两组：正常对照组(17人/17只眼)及原发性开角型青光眼组(17人/17只眼)。受试者在安静坐位休息十分钟期间完成国际身体活动问卷填写工作(International physical activity questionnaire, IPAQ),静坐十分钟后完成运动前眼内压(intraocular pressure, IOP),收缩期血压(systolic blood pressure,SBP),舒张期血压(diastolic blood pressure, DBP),心率(heart rate, HR),视野及脉络膜图像信号采集工作；进行20mmin慢跑有氧运动(每4分钟定时监测血氧饱和度及心率变化)；运动后依次进行眼压,血压,脉络膜图像信号采集工作。利用Image J软件(version1.42, National Institutes of Health, Bethesda, Maryland,USA)提取脉络膜厚度信息(choroidal thickness, ChT)。通过公式计算国际身体活动问卷评分(IPAQ score),最大心率百分比(HR%max),平均动脉压(meanaterial pressure, MAP)及眼灌注压数值(ocular perfusion pressure, OPP)。运用独立样本T检验,独立样本配对T检验,箱图统计学分析方法及受试者工作特征曲线分析(receiver oprating characteristic curve, ROC曲线)对运动前后眼压,平均动脉压,眼灌注压及脉络膜厚度进行相关统计学分析。结果：1.有氧运动后对照组及POAG组患者脉络膜厚度均显著降低,差异具有统计学意义(P0.05)；2.有氧运动后对照组及POAG组患者眼压均显著降低(P0.05)；平均动脉压及眼灌注压均显著升高(P0.05);3. POAG组运动后眼压,平均动脉压及脉络膜厚度波动幅度均与对照组存在统计学差异(P0.05)； 4.受试者工作曲线提示运动前后脉络膜厚度变化曲线下面积(areas under curve,AUC)为71.9%,眼压变化AUC为21.8%。脉络膜厚度变化在区分正常人及POAG患者的过程中具有中等程度分辨作用。结论：1.运动后交感神经系统兴奋性升高可能是导致脉络膜厚度降低,眼压降低及眼灌注压升高的主要原因。2. POAG患者自主神经病变可能是导致其运动后眼压,平均动脉压及脉络膜厚度波动幅度均与对照组存在显著差异的原因。3.本研究采用中等强度运动,研究结果显示,正常人主要通过提高平均动脉压增加眼部血流灌注,而运动后POAG患者主要通过降低眼内压增加眼部血流灌注。故课题组提出以下假设：在日常生活中(低,中强度运动),眼内压持续性波动幅度较正常人增加可能是部分POAG患者视功能损害不断进展的危险因素之一。如何为不同青光眼患者确定合适的运动量可能是下一步需要研究的新方向。
[Abstract]:From the cognitive level, the target information is excavated, the signal is collected by the existing hardware system, the pickup signal is processed and analyzed through the software system and the specific mathematical algorithm, and the target signal is accurately extracted in the interference noise. It is called the Tanner WP first proposed for the signal detection and analysis in.1954.
The eyeball, as an extension of the central nervous system, constantly transforms the received optical signals into electrical signals under stable intraocular pressure and provides abundant visual stimuli for the organisms. How to accurately collect, handle and analyze useful eye signals by using the existing theoretical cognition and using hardware and software systems, and to analyze and analyze useful eye signals, is an eye letter. The core and difficulty in the practical application of the theory of number detection and analysis.
This study is divided into four main parts: (1.) under the theoretical support of the existing clinical detection methods for the pickup process of the diurnal intraocular pressure fluctuation signal, the first part mainly through the medical cross, using the principle of mechanical measurement, to establish a new method of collecting, extracting and analyzing the wave signal of the intraocular pressure; 2. based on the abnormal direct eye blood flow. Under the theoretical support of the progress of optic nerve disease, the second, third and four parts mainly use the existing non-invasive detection techniques and corresponding instruments (coherent optical tomography) to extract the normal people, ametropia and glaucoma patients' eye net membrane and choroid signal after selecting the best measurement techniques and corresponding instruments (coherent optical tomography). The relationship between weave and retinal degeneration, and the difference in response to choroidal signals between normal subjects and patients with primary open angle glaucoma before and after aerobic exercise.
First, the principle and method of continuous intraocular pressure measurement based on magnetic repulsion force.
Objective: to extract intraocular pressure signals from the principle of "permanent magnet magnetic repulsion force measurement" is expected to achieve the ultimate goal of 24h intraocular pressure monitoring.
Principle: the distance between the permanent magnet is gradually reduced and the magnetic repulsion gradually increases. When the magnetic repulsion (the repulsive force between the force magnetic ring and the force magnetic ring) increases and the intraocular pressure is balanced, the magnetic distance will continue to be shortened, and the cornea will be deformed under the external force. At this time, the small amplitude fluctuation can be observed in the electric pressure displacement curve obtained by the oscilloscope. The pressure output signal (U) is related to the intraocular pressure (P). By establishing the mathematical model of the corresponding relationship between U and P, the U signal is extracted from the signal of the pressure curve, and the size of the intraocular pressure P signal can be obtained indirectly.
Methods: using 35mmHg as the upper limit of IOP signal detection, the size and displacement range of permanent magnet magnetic repulsive force needed to cause corneal deformation was evaluated by "finite element simulation method". A simulated eyeball perfusion model was established, and a mechanical measuring device was built to extract the voltage output signal of sensor under different pressure.
Results: 1.N40H rare earth permanent magnet materials can meet the needs of this study, producing sufficient magnetic repulsion force to oppress the cornea and the voltage displacement curve of.2. under different pressures. The output voltage signal of the sensor is different at the fluctuation point of the voltage displacement curve under different pressure. With the gradual increase of the simulated intraocular pressure (10mmHg, 30mmHg, 50mmHg), the output voltage signal of the pressure sensor is raised correspondingly. The two has a good correlation.
Conclusion: this method can indirectly reflect the fluctuation of intraocular pressure by picking up the voltage output signal of capacitive baroreceptor. The overall measurement process is noninvasive and fast. It is expected to achieve the ultimate goal of non-invasive 24h intraocular pressure monitoring.
Two, study on the detection system of eyelid intraocular pressure
Objective: to extract the intraocular pressure signal from the "principle of elasticity difference in biological tissue", which is expected to achieve the ultimate goal of user self-monitoring.
Principle: "the eyelid intraocular pressure detection system" is based on the differences in the elastic mechanics of biological tissues. The IOP signal is obtained by using the "force deformation" difference between the different eyeball tissues.
Methods: the study adopts the mechanical mechanical "compression measurement" to compress the eyelid and the eyeball wall (corneal tissue) and obtain the pressure reflection signal through a capacitive pressure sensor. The joint displacement sensor (variable resistor) collects and draws the voltage displacement curve by the oscilloscope, and the curve "turning point" represents the different tissue junction. The slope of the curve after the inflection point is related to the intraocular pressure. The slope of the curve after the inflection point is calculated, and the correlation between the slope of the curve and the intraocular pressure is observed, and the preliminary research work of the system is completed.
Results: 1. the range of intraocular pressure (13-33mmHg) was measured in this study. The amplitude.2. of intraocular pressure in most patients with primary open angle glaucoma could be measured repeatedly under different perfusion pressures and 4 times. The results showed that the repeatability of the voltage displacement curve was good.3. with the intraocular pressure, and the slope of the voltage displacement curve increased after the inflection point. After fitting, the amplitude of intraocular pressure can be indirectly obtained by the slope of voltage displacement curve K.
Conclusion: the principle of the eyelid intraocular pressure measurement system is feasible, the result is good repeatability, it can reflect the changes of intraocular pressure in real time, and can achieve the ultimate aim of user's self measurement of intraocular pressure.
The second part: the difference of extracting choroidal thickness by different methods.
Objective: to recruit normal subjects as the research object, use the coherent optical tomography to extract the choroidal tissue signal of human eyes, and select the Heidelberg measurement software and ImageJ software to measure the choroid thickness, and analyze the repeatability and consistency of the different measurement methods, and choose the optimal measurement. The method laid the foundation for the follow-up study.
Methods: 36 normal subjects were included in this study. The choroidal tissue images were collected by coherent optical tomography in the 6mm range of the central fovea and parabparal concave. The choroidal thickness was measured by Heidelberg eye explore software (version 5.3.3.0, Heidelberg, Germany) and Image J software (version1.42, USA), using Bl. And-Altmann statistical method was used to analyze and compare the consistency and repeatability of the results of different measurement methods.
Results: 1. Image J software measurement repetition coefficient (39.9186) is higher than Heidelberg measurement software (27.3525); 2. two methods of measurement results consistency analysis shows that 95% confidence interval is -18.437 to 63.949um; 3.95% confidence interval accuracy analysis shows that the upper limit range is 16.102 to 111.796um; the lower limit range -66.29 to 21.41umm.
Conclusion: the repeatability and consistency of the Image J measurement software is higher than that of the Heidelberg software. It should be one of the available software for the researchers to measure the choroidal thickness in the future.
The third part: the changes of choroid thickness and photoreceptor cell thickness in myopic patients.
Objective: to extract the choroidal and retinal tissue signals by using coherent optical tomography to find the relationship between the choroidal and retinal tissue degeneration by recruiting normal people and myopic patients as the research population.
Methods: 64 subjects were divided into three groups according to the diopter state: group I: Group (+1.0D to -1.OD); group II: mild to moderate myopia group (-1.0D ~ -6.0D); group III: high myopia group (-6.0D). Use frequency domain coherent optical tomography (spectral domain optic coherence tomography, SD-OCT) to extract the image of choroid and retina tissue Image J software is used to extract the retinal thickness (Retinal thickness, RT), retinal nerve fiber layer thickness (retinal nerve fiber layer thickness, RNFLT), retinal ganglion cell layer thickness (ganglion), retinal photosensitive cell layer thickness, and choroid thickness Oroidal thickness, ChT) information.
Results: the thickness of photoreceptor layer, the thickness of retina and the thickness of choroid were significantly decreased (P0.05) in the patients with high myopia. The single factor and multi factor linear regression analysis suggested that the changes of the thickness of the photoreceptor layer were closely related to the changes of the choroid thickness.
Conclusion: not only the thickness of the choroid in the patients with high myopia is lower, but the thickness of the photoreceptor layer is also reduced. The two groups constitute the neurovascular unit. The change of the choroidal thickness will lead to the change of the thickness of the photoreceptor layer, and vice versa.
Objective: in this study, the patients with primary open angle glaucoma (primary open angle glaucoma (POAG)) were used as the main research population, using the coherent optical tomography (optic coherence tomography, OCT) to extract the image signal of the choroid tissue of the subjects, and to extract the intraocular pressure signal by non-contact tonometer, and to explore the preoperative and the aerobic exercise. Changes of ocular choroidal thickness and intraocular pressure in normal subjects and POAG patients after aerobic exercise.
Methods: 34 subjects were divided into two groups: the normal control group (17 /17 eyes) and the primary open angle glaucoma group (17 /17 eyes). The subjects completed the international physical activity questionnaire (International physical activity questionnaire, IPAQ) during the rest period of rest ten minutes, and completed the pre exercise eye after ten minutes of sitting. Pressure (intraocular pressure, IOP), systolic blood pressure (systolic blood pressure, SBP), diastolic blood pressure (diastolic blood pressure, DBP), heart rate (heart), visual field and choroidal image signal acquisition, and aerobic exercise (monitoring blood oxygen saturation and heart rate changes every 4 minutes); after exercise Image J software (version1.42, National Institutes of Health, Bethesda, Maryland, USA) extraction of choroidal thickness information (choroidal thickness). The formula is used to calculate the international body activity questionnaire, the maximum heart rate percentage, and the mean arterial pressure. Naterial pressure, MAP) and eye perfusion pressure value (ocular perfusion pressure, OPP). Independent sample T test, independent sample paired T test, box graph statistical analysis method and subject work characteristic curve analysis (receiver oprating characteristic), intraocular pressure, mean arterial pressure, eye perfusion pressure and pulse before and after exercise The thickness of the collaterals was analyzed statistically.
Results: 1. after aerobic exercise, the choroidal thickness of the control group and the POAG group decreased significantly, and the difference was statistically significant (P0.05); 2. after aerobic exercise, the intraocular pressure of the control group and the POAG group were significantly decreased (P0.05); the mean arterial pressure and the ocular perfusion pressure were significantly increased (P0.05); the intraocular pressure after exercise, the mean arterial pressure and the choroid membrane in the 3. POAG group were observed. There was a significant difference in thickness fluctuation between the control group and the control group (P0.05).
The work curve of 4. subjects suggested that the area under the curve of choroidal thickness (areas under curve, AUC) before and after the movement was 71.9%, and the change of intraocular pressure AUC to 21.8%. choroid thickness was of moderate resolution in the process of distinguishing between normal and POAG patients.
Conclusion: 1. the increased excitability of the sympathetic nervous system after exercise may be the main cause of the decrease of choroidal thickness, the decrease of intraocular pressure and the increase of intraocular perfusion pressure in.2. POAG patients, which may be the cause of the intraocular pressure, the mean arterial pressure and the amplitude of the choroidal thickness are significantly different from those in the control group.3. The study uses moderate intensity exercise. The results show that normal people increase the blood flow of the eye mainly by increasing the mean arterial pressure, and the POAG patients mainly increase ocular blood flow by lowering intraocular pressure after exercise. Therefore, the group proposes the following hypothesis: in daily life (low, moderate intensity exercise), the constant fluctuation range of intraocular pressure is more than that of the normal people. The increase in normal people may be one of the risk factors for progressive visual impairment in some POAG patients. How to determine the appropriate amount of exercise for different glaucoma patients may be a new direction in the next step.
【学位授予单位】：华中科技大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：R770.4

【共引文献】