Instrumentation for Automated Study of Human Pupil Responses

Instrumentation has been developed for studying the responses of the human pupil to different kinds of visual stimuli. This involves stimulating the eye visually, acquiring pupillary diameter data samples, analyzing them, and displaying and storing the results All of these functions are controlled by a minicomputer. The apparatus is a convenient research tool for studying the visual system of color-normals and color-defectives by observing pupillary behavior.     

A Distributed Model of Iron Kinetics for Clinical Assessment of Normal?Abnormal Erythropoietic Activity

The use of mathematical models of iron metabolism has been recognized as an essential tool for analyzing ferrokinetic data. Multicompartment models, however, did not make it possible to describe marrow iron kinetics, taking into account that marrow erythroid cells uptake iron from the plasma according to their maturity.     

基于OpenCV的人眼瞳孔定位算法

近年来,人脸检测技术在机器视觉、模式识别、人工智能等领域得到了广泛的应用。而包含着丰富的内容和信息的人眼,作为人体生物特征的一个重要部分,关于人眼检测与定位的技术更是广泛应用在生物识别、医疗辅助、人机交互等领域。该文利用水平积分投影和区域特征的算法,快速定位人眼区域,然后利用区域投影的方法实现了瞳孔定位。     

正常和异常血清吸收光谱的实验研究

生命科学是当今科学研究的一个重要领域,本文针对血液检测这一研究课题,采用日本岛津紫外-可见近红外分光光度计研究了正常血清、高血糖血清、高血脂血清的吸收光谱的异同,并探讨了正常血清与异常血清在某些波段上吸光度的不同,发现正常血清和高血糖血清的吸收光谱没有太大差别,不能用吸收光谱来判断血清中的血糖是否超标;高血脂血清和正常血清相比,随着血脂浓度的提高,血清的吸光度增大.尤其是在416nm处的吸光度,提出了利用血清在某些波段上的吸收率来判断血清中血脂是否异常的方法,为医学应用提供了新的分析方法.     

优化ALA-PDT对正常人角质形成细胞凋亡的影响

目的:本文通过不同浓度的ALA与不同用药时间对体外培养的人角质形成细胞(KC)PDT的生物效应研究,观察PpⅨ荧光强度及KC凋亡和生长周期的影响,优化ALA最佳药物浓度和用药时间,探讨ALA-PDT抑制银屑病KC异常增生的可行性和最佳效果.方法:将新鲜包皮组织经两次酶消化法进行KC分离与培养,分设KC对照组、单纯ALA组、单纯照光组及0.1、0.6、1.2、1.8、3.6mmol/L ALA五个浓度组,经0.5、1、3、5h四个避光孵育时间后PDT.FCM、酶标仪、荧光显微镜检测KC中PpⅨ荧光强度,确定ALA最佳药物浓度和最佳用药时间;Hoechst33342 / PI双染法和Annexin V / PI双染法检测KC凋亡率和对生长周期的影响.结果: 0.6mmol/L ALA(用药1h)-PDT组为最佳药物浓度和最佳用药时间, 显示PpⅨ荧光强度表达与KC凋亡率最高,能明显抑制S期与G2期的细胞增值,使细胞增殖指数降为最低,与其他各组相比差异显著(P<0.05).结论: 0.6mmol/L ALA(用药1h)-PDT能明显抑制KC增殖,优化后的ALA-PDT更迅速有效地促进人KC凋亡.     

基于计算机视觉的人体异常行为识别综述

人体行为识别是计算机视觉和模式识别领域的研究热点之一。作为人体行为识别的一个重要分支,人体异常行为检测近年来也不断得到学界及工业界的重视。人体行为识别研究从早期的依赖人体形状特征发展到基于梯度设计的特征检测,再到当前随着神经网络的新发展,深度学习开始广泛应用于行为识别。同时由于红外波段具有适应弱光照环境、可全天候检测等优点,基于该波段的人体行为识别研究开始兴起,它也必将成为人体行为识别领域中一个新的研究热点。     

基于粒子流和能量模型的异常行为检测

提出了一种基于单目摄像机的异常行为检测方法,可有效检测群聚群散事件。针对传统稠密光流法耗时大,抗噪声性差,对光照变化敏感,而稀疏光流又不能提供丰富运动信息的不足,提出采用二值前景为掩码计算局部稠密光流,既剔除了干扰,又减少耗时。针对光流仅能反应物体瞬时运动信息,不能反应运动的累积效应,不能描述物体间的相互影响,提出建立以光流为底层特征的粒子流场,以稳定地进行特征提取。最后,通过计算粒子幅值直方图的能量来进行异常行为的检测。搭建了实验监控系统,不同场景的实验表明,该方法实时性好,鲁棒。     

A Statistical Analysis of Pupil Noise

The random fluctuations of human pupillary area provide an interesting example of a stochastic process in a biological system. Statistical theory and system analysis have guided the authors in performing experiments which reveal important properties involved in the random process.     

基于多条件随机场模型的异常行为检测

传统的异常行为检测多数是利用单特征建模,检测的行为较为单一,检测率较低,针对这些问题,提出一种基于多条件随机场模型（MCRF）的异常行为检测方法,MCRF模型具有融合多特征和联系上下文信息的优势。通过Kinect获取3D骨架数据,提取角度、位置、速度三类特征,形成多类特征子集,利用基本的CRF模型对每一类特征子集建模,形成多个CRF单元,然后组合所有的CRF单元,得到MCRF模型,最后利用MCRF模型进行异常行为检测。实验结果表明基于MCRF的异常行为检测方法具有较高的检测率。     

Nuclear Morphometry in Normal and SV-40 Transformed Human Diploid Fibroblasts

These experiments were designed to compare the densitometric, morphometric, and texture characteristics of intact nuclei from stationary WI-38 human diploid fibroblasts and their stationary SV40-transformed counterparts (2RA cells) by means of automated image analysis and to compare the differences in template activity of the chromatins isolated from the two cell lines. While the frequency distributions of directly measured parameters as area and IOD show a substantial overlap between WI-38 and 2RA cells, average optical density parameter derived as IOD per unit area, yields two distinctly separated distributions, which permit objective identifications of individual SV40 transformed cells. Chromatin of the stationary 2RA cells appears more condensed than that of the confluent WI-38 cells, quite compatible with the decreased template activity of chromatin from transformed cells.     

人眼光学模型的研究与发展

以适宜光刺激为主的人眼是人类获取外界信息最多的重要感觉器官.从光学角度而言,人眼是一个有多个界面的复杂光学系统,每一个界面都具有复杂的几何特征和光学特性.结合光学测量,对人眼建立光学模型,研究每个界面的光学特征,在视光学和眼科学领域都有着重要的实际应用意义.通过对国内外模型眼的研究与发展的介绍,总结了各种典型人眼模型的光学结构与特性,并给出了人眼光学模型的研究发展趋势--个性化与精密性.     

正常人周围神经冷冻复型超微结构观察

本研究利用冷冻复型电镜技术对５例正常人周围神经进行了观察：（１）有髓神经纤维髓鞘以同心圆形式包裹轴索，后者位于中央呈圆柱样。（２）雪旺细胞核膜劈裂面不论ＰＦ面还是ＥＦ面膜蛋白颗粒（ＩＭＰ）均较丰富，核孔数量较多，说明该细胞代谢功能较旺盛；另外其胞膜表面ＩＭＰ呈簇状分布，推断在正常情况下该区域物质代谢较活跃或雪旺细胞具有分泌功能；再者，髓鞘板层与一般生物膜不同，其劈裂面不论是ＰＦ面还是ＥＦ面均罕见ＩＭＰ，推测此为神经冲动呈跳跃式传导和绝缘性传导的物质基础。（３）轴索内亚细胞结构可见微丝、微管，二者除具有支架作用外，还参与轴浆物质运转等许多重要的神经生理活动。（４）沿神经纤维纵向劈裂面可见相邻雪旺细胞膜内陷相贴，形成轴索外系膜。     

Gaitspert: An Expert System for the Evaluation of Abnormal Human Locomotion Arising from Stroke

This paper describes the implementation and use of a knowledge-based system for the evaluation of abnormal human locomotion arising from cerebral vascular accidents (CVA). An expert system (GAITSPERT) that is knowledgeable about the evaluation of human gait has been implemented primarily as a clinical tool that can produce assessments of underlying stroke disease states and recommendations about therapeutic interventions. GAITSPERT combines artificial intelligence (AI) analysis methodologies with biomedical signal acquisition and evaluation providing intelligent processing of kinematic, electromyographic (EMG), and foot-switch data. GAITSPERT was constructed using GENIE, a general-purpose knowledge-enginering tool, and a knowledge base of stroke-related facts about neuromuscular disabilities.     

基于正态云模型的网络入侵检测

将正态云模型用于网络入侵检测,并提出一种基于正态云模型的网络入侵检测算法,算法首先采取特征提取并计算影响权值因子,引入云发生器计算属性特征值和差异度,入侵判断采用综合评价值,减少了单个属性造成的局部影响.实验结果表明,该方法能够比较有效地检测真实网络数据中的未知入侵行为,在一定程度上解决了目前部分入侵检测算法存在的检测率低、误报率高的问题.     

基于时空LBP加权社会力模型的人群异常检测

针对基于传统社会力模型的人群异常行为检测算法忽视了场景中各个区域信息差异性的问题,考虑到时空LBP序列谱特征的计算简单性和区域代表性,提出了一种基于时空LBP加权社会力模型的人群异常行为检测算法,将时空LBP序列谱特征所包含的时域特性和区域信息融入社会力模型,使得社会力模型更为精确地对人群行为进行建模。实验证明,与传统算法相比,改进后的算法在异常行为的查准率与查全率上有很大的提高。     

Computer-Aided Medical Instruction Using an Interactive Graphics Model of the Normal and Congenitally Defective Heart

The on-line interactive-graphics digital-computer model of the normal and congenitally defective heart was developed as a basic research and teaching tool. The interactive graphics capability of the simulation system provides the user great flexibility in selecting the types of defects he wishes to study, and in allowing him to progress at his own speed.     

微电子制造业人才需求与培养模式研究

人才短缺、质量不高和层次结构不合理等问题成为制约我国微电子制造业发展的重要瓶颈之一,其产生原因是微电子制造业人才需求与培养存在严重脱节.通过对台积电、中芯国际等微电子制造企业进行深入访谈调研,结合我国微电子制造业发展现状和趋势,提出一种互动式的,高校、企业和政府共同参与的微电子制造业人才需求与培养新模式,并探讨了新模式下的动态平衡及运行机制.     

Linear Homeomorphic Model for Human Movement

The parameter values for this model are specific to the human eye movement systems; however, the form of the model is applicable to other neurological motor control systems. The muscle length-tension diagram was modeled with an ideal spring. The muscle force-velocity relationship was linearized in a manner that produced a linear model. Initial parameter estimates were based on physiological data, human when possible. Then a function minimization program was used to fine tune model parameters. These parameter values were compared to the original physiological data to ensure that they were within the range of variability of the data. The antagonist dashpot value was selected to minimize the mean squared error between human and model responses; the value produced suggested a unique simplified representation for the original physiological data. The parameter estimation routine was applied to make the model match atypical human eye movements; these simulations suggested that glissades in normals are caused by pulsewidth, not pulse height errors.     

Identification of Thermal Model for Human Tissue

A simple model structure of spatial and temporal temperature distribution in human tissue, exposed to electromagnetic heating and surface cooling, is developed. This model, denoted the control model, is useful for thermal dosimetry in hyperthermia cancer therapy. The theoretical control model parameter values for various types of homogeneous tissue are determined from a priori knowledge about tissue composition and parameters. The control model parameters for real tissue are estimated from patient treatment data through the use of a system identification technique. The results indicate that the control model structure is adequate. With the available a priori knowledge the theoretically determined control model parameter values do not accurately reproduce the experimentally estimated values. Accordingly, an identification based on experimental data is recommended, if an accurate model is required.