神经网络在卵巢浆液性囊腺肿瘤细胞核形态分析中的应用

目的：本研究将计算机技术与病理学专家的实际经验相结合，在采用图像处理技术对医学图像进行处理的基础上，应用人工神经网络分析卵巢浆液性囊腺肿瘤细胞核形态特征，并比较径向基函数网络与自组织网络两种算法。方法：对每个良性，交界性，或恶性卵巢浆液性囊腺肿瘤细胞及正常卵巢上皮组织细胞分别提取五个形态参数，用两种神经网络进行分类识别。结果：径向基函数训练步数少，但准确率没有自组织网络高。结论：应用神经网络进行细胞识别。在医学科研以及临床诊断方面有重要的现实意义和广阔的应用前景。     

基于Curvelet变换的肺结节CT图像良恶性分类研究

目的早期肺癌患者的CT图像表现为结节状（在肺野内直径≤3cm的病灶），需要与结核球等良性病变鉴别开，以提高患者的5年生存率。方法本文基于Curvelet变换提取能量、熵、灰度均值及灰度标准差四种纹理特征值，按7：3比例将样本分为训练集与验证集。使用BP（back propagation）神经网络作为分类器。每一种纹理参数测试集的神经网络仿真值结合病理诊断结果绘制受试者工作特征曲线（receiver operator characteristic cllrve，ROC曲线），根据ROC下面积得到最优的几种纹理参数用于良恶性分类，并将分类结果与病理诊断结果进行比较。结果四种纹理参数构建的BP网络均具有诊断价值，每种纹理参数诊断价值各不相同，其中熵与灰度标准差的诊断价值优于能量与灰度均值，并且通过组合多种纹理参数可以提高诊断准确性。结论使用熵与灰度标准差两种纹理特征值构建BP神经网络能达到最好的分类效果，在一定程度上有利于肺癌的早期诊断。     

神经网络对卵巢粘液性囊腺肿瘤细胞识别的研究

目的：探讨人工神经网络在识别卵巢粘液性囊腺肿瘤细胞的价值。方法：使用MATALB软件中的神经网络工具箱（Neural Network Toolbox）,分别设计两种神经网络并训练。网络以五个形态参数（细胞核面积、周长、最大直径、等效圆直径、似圆度）为输入,输出为正常、良性、交界和恶性四种形态类型。结果：训练好的神经网络可以准确的对卵巢肿瘤细胞形态分类。结论：人工神经网络对卵巢粘液性囊腺肿瘤的病理学鉴别有很好的应用前景;选择适合的网络,不仅看网络自身,还要看训练样本。     

红细胞形态特征统计分布及计算机图像分析方法研究

基于计算机图像处理技术,对红细胞形态和图像特征进行分析,建立红细胞形态变化特征分布统计模型,实现红细胞形态特征的自动统计分类。引用形状因子对红细胞形态特征进行描述和分类,建立红细胞分割、识别和测量的相关模型和算法,并用VC 引进编程实现。该方法能快速自动地对红细胞形态特征作出统计分析,结果显示溶血性贫血症的红细胞形态分布（双峰曲线）与正常红细胞形态分布（单峰曲线）有显著差异,即统计分布曲线可获得不同形态和类型的红细胞比例的数据。在以正常红细胞形态特征参比的条件下,红细胞形态特征统计方法可用于与红细胞形态变化相关疾病（如溶血性贫血、早期白血症）的辅助诊断。     

小学生(四-六年级)学习能力倾向测验的初步编制Ⅱ:效度检验

目的：验证小学生（四—六年级）学习能力倾向测验的效度。方法：采用探索性因素分析和验证性因素分析研究测验的结构效度：对测验分数与教师的评定等级进行相关分析，比较样本中来自不同等级学校、不同年级，或性别不同的学生之间测验得分的差异，以检验同时效度：以回归分析检验测验分数对语文、数学考试成绩的预测效度。结果：①探索性因素分析和验证性因素分析均证实了原先设想的测验结构；②语文分量表的分数与教师评定的语文学习能力等级相关为0．53（A卷）或0．58（B卷），数理分量表分数与教师评定的数学能力等级相关为0．58（A卷）或0．56（B卷），空间分量表分数与数学能力等级相关为0．25（A卷）或0．3l（B卷）；各个分测验分数大体上均呈随年级而逐步递增的趋势；各分测验的得分呈现优等学校〉中等学校〉差等学校的趋势；女生在某些语文分测验上的分数明显优于男生（P均〈0．05或0．01）；（爹第一主成分回归的决定系数分别为：语文0．471（A卷）或0．386（B卷），数学0．491（A卷）或0．362（B卷）（P均〈0．001）。结论：测验具有良好的结构效度、同时效度和预测效度。     

人脑松果体区域的断面解剖学研究

应用断面解剖学方法，对３３例（男１８，女１５）头颅标本的松果体及其眦邻结构进行了观测。结果发现：有８７．８８％的松果体中点偏离中线，偏离中线的平均距离及与之对应的移位百分率是：左移０．９９ｍｍ（０．８０％），右移０．７１ｍｍ（０．５４％）；松果体及第三脑室长径、宽径均无性差，经松果体中点半球宽存在侧差（左侧〈右侧）和性差（男〉女），左、右半球长径存在性差（男〉女）；此外，本文应用逐步回归方法对松果     

仫佬族地区护士群体心理健康及其与人格特征、应对方式的相关性研究

目的调查仫佬族地区护理群体心理健康及其与人格特征、应对方式的相关性。方法采用分层随机抽样的方法对152名护理人员进行SCL-90、EPQ、SCSQ测评。结果仫佬族地区护理人员强迫、恐怖、精神病性因子分均高于中国常模（P〈0.05或P〈0.01）。护士的强迫、恐怖、精神病性症状均与精神质、消极应对呈正相关（P〈0.01），与内外向呈负相关（P〈0.05或P〈0.01）；强迫、精神病性症状与神经质呈正相关（P〈0.05）；3个症状均与积极应对无关（P〉0.05），恐怖与神经质无相关性（P〉0.05）。不同年龄段护士SCL-90评分比较，差异无统计学意义（P〉0.05）。结论仫佬族地区护士群体心理健康水平与个体特有的个性心理特征有关；积极的应对方式有利于心理健康，消极的应对方式不利于心理健康。     

精子密度、存活率、活动力、形态与宫腔内人工授精妊娠结局

目的探讨精子密度、存活率、活动力（a＋b），形态与宫腔内人工授精妊娠结局的相关性。方法选自以男性不育为因素289对夫妇进行IUI周期治疗：在授精日，取男方精液进行计算机综合分析，常规方法授精后，将其精液检测结果与妊娠结局相比较。结果精子密度、存活率、活力（a＋b）％、形态与IUI妊娠结局概率分别是：15．2％，16．3％，16．9％，18．7％；精子密度组：（30～60）×10^6／ml与（60～90）X10^6／ml、≥100×10^6／ml、≤20X10^6／ml比较有差异性P〈0．01；存活率组：30％-60％与60％～80％、≥80％比较有差异性P〈0．01；活力组：30％～45％与10％～20％、20％～30％、〉50％组间比较有差异性P〈0．Ol；形态组：〈15％纽与a〉30％与30％≥A≥15％比较有差异性P〈0.01。结论精子密度、存活率、活力（a＋b）％、形态与IUI妊娠结局密切相关，可以作为评价宫腔内人工授精的一项实验室指标；精子形态与IUI妊娠结局呈正相关；精子较低{密度、活率、活力（a＋b）％}与IUI妊娠结局呈正相关；单纯性男性少、弱精子症在IUI周期中能获得较高的妊娠率。     

抑郁症治疗前后脑白质弥散张量成像研究

目的：探讨中青年重性抑郁症患者治疗前后的脑白质纤维及认知功能变化。方法：对11例重性抑郁症患者分别于治疗前和治疗后10周评定24项汉密尔顿量表（HAMD24）、威斯康星卡片分类测验（WCST）、韦氏记忆量表及弥散张量成像（Diffusion Tensor Imaging，DTI）扫描，并与11例健康对照比较。结果：（1）抑郁组治疗后的WCST成绩（完成分类数：4.0±2.1、应答错误数：43．2±18．8、持续性错误数：22.8±16.0）较治疗前明显改善（1.6±1.6、67．9±20.0、51．5±24．8，P=0．001、0．000、0．003），而瞬时记忆力（顺背及倒背）治疗前后差异均无显著性（P〉0．05）；HAMD总分治疗后低于治疗前（16±14／54±13，t=6．60，P〈0．001）；治疗后抑郁组与健康对照组WCST和记忆力得分差异均无显著性（P〉0．05）；（2）治疗后抑郁组在右额叶下（26；13；-11，52像素）及左胼胝体区（-8；26；-1，48像素）白质的各向异性（fractional anisotropy，FA）值较疗前升高，但同健康对照组相比，左颞叶下（-32；-52；15，34像素）、右扣带回（18；-25；28，118像素）及右脑岛叶（32；-4；17，39像素）区域白质的FA值仍存在一定程度的降低（P〈0．001，像素〉30）。结论：抑郁症患者治疗后认知功能明显改善；治疗后脑白质纤维亦有一定的恢复，但同正常人相比，仍存在一定程度的白质损害。     

CD30阳性的间变性大细胞淋巴瘤临床病理研究

目的；研究ＣＤ３０阳性的间变性大细胞淋巴瘤（ＡＬＣＬ）的病理形态和免疫一型，并与多种恶性肿瘤相鉴别，方法：对９例ＡＬＣＬ淋巴结活检材料常规处理后，ＨＥ染色及免疫组化染色（ＡＢＣ法），结果：９匀呈ＣＤ３０阳性，４例为Ｔ细胞怀，３例为Ｂ细胞性，２例既无Ｔ也无Ｂ细胞免疫表型表达。其形态特征为：部分或全部淋巴结受累。肿瘤细胞核为多形，窦内浸润及巢状浸润。结论：ＡＬＣＬ为非同源量形态特征相似。需与以下疾病鉴     

Design and testing of a genetic algorithm neural network in the assessment of gait patterns

It is important to be able to quantify changes in gait pattern accurately in order to understand the clinical implications of surgery or rehabilitation. Although supervised feed-forward backpropagation neural networks are very efficient in many pattern-recognition tasks, the genetic algorithm neural network (GANN), which can search in some appropriate space, has not been used previously for gait-pattern recognition. This study discusses how to use the GANN approach in gait-pattern recognition, and evaluates the complexity and training strategy of the particular classification problem. Both the GANN and a traditional artificial neural network (ANN) were used to classify the gait patterns of patients with ankle arthrodesis and normal subjects. The GANN model was able to classify subjects with recognition rates of up to 98.7%. In contrast, the ANN trained by using all possible predictor variables was only able to classify the subjects with recognition rates of 89.7%. It is suggested that the GANN model is more suitable to exploit the patient's gait pattern. The value of the neuron output can be used as an index of the difference from normal. By this means, all pathological gait patterns may be presented quantitatively.     

利用Stein-Weiss解析函数结合反向传播神经网络进行血管分割

针对传统反向传播（BP）神经网络对血管进行分割存在耗时长且识别率不高的问题,本研究提出一种新的基于Stein-Weiss解析函数的BP神经网络算法用于血管分割。首先为每个体素构建一个Stein-Weiss函数,然后根据Stein-Weiss解析函数的解析性,计算出相应体素的16个特征值,将这些特征值输入到BP神经网络的输入层,采用BP神经网络的自学习能力对这些数据进行分类学习,最后通过BP神经网络的泛化能力来获取血管边缘。对肝脏血管分割的实验结果表明,相对于传统的BP神经网络分割算法,该算法提取的函数血管边缘识别率高、细节丰富,分割效率也明显提高。     

神经网络在染色体自动解析系统中的应用

神经网络是近年来发展起来的新型计算系统 ,广泛应用于模式识别、特征提取和信号处理等方面。文章论述了目前国外染色体自动解析系统的设计发展现状 ,主要讨论了神经网络在这一方面的应用。文章从染色体的结构、神经网络的基本原理讲起 ,涉及到国外众多学者的研究与实验 ,实际反映了基于神经网络的染色体自动解析系统的国际发展水平 ,并指出了系统设计中存在的缺陷。     

The application of neural networks in predicting the outcome of in- vitro fertilization

Infertility affects one in six couples at some time in their lives, with 48% of these couples requiring assisted conception techniques in order to achieve a pregnancy. Whilst the overall clinical pregnancy rate per embryo transfer is 23%, this varies widely between clinics. The Human Fertilisation and Embryology Authority has attempted to analyse the results of all units, with weighting of different factors affecting assisted conception, and the published data have invariably led to comparisons between units. However, statistical models need to be developed to eliminate bias for valid comparisons. Neural networks offer a novel approach to pattern recognition. In some instances neural networks can identify a wider range of associations than other statistical techniques due in part to their ability to recognize highly non-linear associations. It was hoped that a neural network approach may be able to predict success for individual couples about to undergo in-vitro fertilization (IVF) treatment. A neural network was constructed using the variables of age, number of eggs recovered, number of embryos transferred and whether there was embryo freezing. Overall the network managed to achieve an accuracy of 59%.      

Simulation studies of data classification by artificial neural networks: Potential applications in medical imaging and decision making

Artificial neural networks are being investigated in the field of medical imaging as a means to facilitate pattern recognition and patient classification. In the work reported here, the effects of internal structure and the nature of input data on the performance of neural networks were investigated systematically using computer-simulated data. Network performance was evaluated quantitatively by means of receiver operating characteristic analysis and compared with the performance of an ideal statistical decision maker. We found that the relatively simple neural networks investigated in this study can perform at the level of an ideal decision maker. These simple networks were also found to learn accurately even when the training data are extremely unbalanced with respect to the prevalence of actually positive cases and to differentiate input data patterns by recognizing their unique characteristics.     

基于支持向量机的肝纤维化CT图像分类

基于统计学习理论中结构风险最小化原则的支持向量机是易于小样本的机器学习方法.本文使用支持向量机和二叉树的方法对肝纤维化CT图像进行分类,并与k近邻法和BP神经网络等其它算法进行比较,结果显示对于肝纤维化图像,支持向量机的分类效果和鲁棒性要高于其他两种算法.     

Modified adaptive resonance theory based control strategy for EMG operated prosthesis for below-elbow amputee

This paper presents a modified adaptive resonance theory (ART1)-based control strategy for a below-elbow (BE) prosthesis. The statistical parameters and histogram from two channels of an electromyogram (EMG) signal have been used as the feature space for the classification of four limb functions. The ART1 neural network (NN) has been used for the classification. ART1 has been modified to learn the patterns in supervised manner to suit the application. Further, the criteria for the modification of the stored pattern have been made bi-directional and the matching criteria have been designed for bit-by-bit matching. The major challenge of using ART1 is to decide on the value of the vigilance parameter, as the classification success is drastically affected by this parameter. The criteria have been evolved to get the optimal value of the vigilance parameter. It is concluded that the best value of vigilance parameter is that which provides the same tolerance in matching as the minimum bit distance between the stored patterns. This scheme has also been implemented on an 8031 microcontroller.     

An exact statistical method for comparing topographic maps,with any number of subjects and electrodes

Summary Statistical methods for testing differences between neural images (e.g., PET, MRI or EEG maps) are problematic because they require (1) an untenable assumption of data sphericity and (2) a high subject to electrode ratio. We propose and demonstrate an exact and distribution-free method of significance testing which avoids the sphericity assumption and may be computed for any combination of electrode and subject numbers. While this procedure is rigorously rooted in permutation test theory, it is intuitively comprehensible. The sensitivity of the permutation test to graded changes in dipole location for systematically varying levels of signal/noise ratio, intersubject variability and number of subjects was demonstrated through a simulation of 70 different conditions, generating 5,000 different data sets for each condition. Data sets were simulated from a homogenous single-shell dipole model. For noise levels commonly encountered in evoked potential studies and for situations where the number of subjects was less than the number of electrodes, the permutation test was very sensitive to a change in dipole location of less than 0.75 cm. This method is especially sensitive to localized changes that would be washed-out by more traditional methods of analysis. It is superior to all previous methods of statistical analysis for comparing topographical maps, because the test is exact, there is no assumption of a multivariate normal distribution or of the correlation structure of the data requiring correction, the test can be tailored to the specific experimental hypotheses rather than allowing the statistical tests to limit the experimental design, and there is no limitation on the number of electrodes that can be simultaneously analyzed.This research was funded in part by the University of South Florida's Research and Creative Scholarship program.     

Modified adaptive resonance theory based control strategy for EMG operated prosthesis for below-elbow amputee

This paper presents a modified adaptive resonance theory (ART1)-based control strategy for a below-elbow (BE) prosthesis. The statistical parameters and histogram from two channels of an electromyogram (EMG) signal have been used as the feature space for the classification of four limb functions. The ART1 neural network (NN) has been used for the classification. ART1 has been modified to learn the patterns in supervised manner to suit the application. Further, the criteria for the modification of the stored pattern have been made bi-directional and the matching criteria have been designed for bit-by-bit matching. The major challenge of using ART1 is to decide on the value of the vigilance parameter, as the classification success is drastically affected by this parameter. The criteria have been evolved to get the optimal value of the vigilance parameter. It is concluded that the best value of vigilance parameter is that which provides the same tolerance in matching as the minimum bit distance between the stored patterns. This scheme has also been implemented on an 8031 microcontroller.     

Prediction of protein functional domains from sequences using artificial neural networks

An artificial neural network (ANN) solution is described for the recognition of domains in protein sequences. A query sequence is first compared to a reference database of domain sequences by use of and the output data, encoded in the form of six parameters, are forwarded to feed-forward artificial neural networks with six input and six hidden units with sigmoidal transfer function. The recognition is based on the distribution of scores precomputed for the known domain groups in a database versus database comparison. Applications to the prediction of function are discussed.