基于人工神经网络的慢性肾病分级辅助诊断专家系统设计

目的：慢性肾病分级辅助诊断系统设计的目的是共享专家经验,使医生更准确的诊断慢性肾病,使慢性肾病患者及时了解自己的病情,同时收集病例数据。将收集到的病例资料建立成数据仓库。评估数据质量,使用决策树方法进行数据挖掘,对数据进行分析。方法：本文以GFR估算模型为基础,设计由客户端、数据库服务器、和网站组成的专家系统。系统采用C／S及B／S混合体系机构,在Windows7的操作平台下,以Qt为开发平台,采用C＋＋编程语言进行开发。结果：医生只需要在PC机运行客户端,就可以进行数据的录入,通过选择GFR估算方法即可得到GFR预测值。实验中,以中山大学附属第三医院。肾内科病例数据为依据,验证系统。系统平稳,能较好的实现人工神经网络算法在慢性肾病分级预测中应用。本系统能够快速准确预测病情分级,并提供规范的治疗方案。数据库实现了对病历资料查询、管理、收集数据等主要功能。结论：本系统实现基于C＋＋语言的GFR估算平台,实现了已有模型的外部扩展使用和数据库对数据管理功能。系统界面友好、操作简单,有效提高慢性肾病诊断确诊时间、分级预测正确率。     

基于医院信息系统的多学科诊治肿瘤数据库系统设计

目的针对肿瘤的多中心诊疗、病历数据繁杂等特点,设计了一种对肿瘤临床数据可进行有效管理、信息共享、查询统计及数据挖掘的数据库系统。方法通过导出医院信息系统（hospital information system,HIS）中所需的信息到中间数据库,运用ASP．NET技术对中间库进一步过滤,将数据整理为具有结构化、标准化字段的肿瘤数据库。结果本系统实现了病历分析、统计,并采用基于模糊数学的方法实现了临床辅助决策。结论该系统可规范医生诊治流程,有利于多科室共同治疗,体现了多学科一体化模式在诊治肿瘤中的医疗价值和社会价值。     

基于超声病历的数据库管理与分析系统

为了解决超声医学技术的快速发展带来的病历数激增而难以管理的问题,完成了超声病历数据库管理与分析系统的设计与验证。系统基于SQL Server数据库技术、Visual C#窗体应用程序技术、医学数据统计分析技术成功构建了一套以超声为核心的病历数据库管理与分析系统,同时系统留有电子病历、实验室检查、病理检查等数据接口,并将病历数据的常用统计分析算法结合起来。通过导入原有病历数据试验,超声科医生对于病历数据库的管理分析及相关课题研究效率提升50%以上。试验结果表明该系统有助于超声科医生对于病历数据的管理与分析,对提高医疗科研效率具有实际意义。     

基于磁共振图像与波谱的专家辅助诊断系统的设计

神经系统疾病是影响人类寿命和生存质量的常见疾病,磁共振图像(MRI)和磁共振波谱(MRS)分别提供了病变结构像和代谢物含量改变信息。综合分析MRI和MRS信息有助于医生更好地诊断疾病和制定治疗方案,因此磁共振数据共享与辅助诊断平台的建设受到广泛关注。基于MRI和MRS的专家辅助诊断系统以脑肿瘤、阿尔茨海默症、轻度认知障碍为重点,以DICOM格式标准的MR图像和LCModel量化后的MRS结果为数据基础,使用MySQL建立后台数据库存储和管理医学数据,并采用Java进行系统的后台和前端开发。该系统提供了病人信息查询、MRI及其属性标签查看与下载、MRS量化结果显示与下载、MRS与MRI的位置配准及其空间可视化等功能,为医生、研究人员、管理员等提供不同权限的需要。该系统的建立为医生的病情诊断和科研人员实施研究提供了极大帮助,推进了远程资料共享和自动化诊断的进程,并促进临床和科研的学术交流。     

基于SQL SERVER2008的眼科电子病历系统的研制

目的：电子病历系统在现代医学信息处理中占有很高的地位,本文针对医院对于眼科病人信息管理的实际需求,开发了一个眼科电子病历系统,以实现对患者的诊疗信息的管理及眼底图像的处理。方法：系统采用SQL SERVER2008作为应用数据库,ActiveX数据对象（Active Data Object简称ADO）的方式连接数据库,以Visual Studioc＋＋6．0作为开发语言,采用SQL语句拼接的方法实现高级查询,采用改进的Back Propagation（BP）神经网络算法进行图像的分割及处理。结果：该眼科电子病历系统包括系统管理、病人信息管理、信息查询及图像管理等四个模块,可实现患者信息的管理、高级查询及统计分析等功能,另外,还可对眼科眼底造影图像进行处理,可手动及自动测量眼底血管管径及不规则病变的面积。结论：通过SQL语句拼接实现对数据库的高级检索,极大地提高了数据库的操作效率。系统给出的眼底血管和病变面积的测量数据,可为眼底病变的诊断、治疗及预后评估等提供依据。总之,该眼科电子病历系统提供了一个操作简便．界面友好的工具,有助于临床诊疗、教学及科研工作。     

医院实验室信息系统的开发和应用

目的为检验科实现自动化数据传输，将检验结果数据及时、准确地回送到医生站、护士站而设计、开发、应用实验室信息管理系统（laboratory information system，LIS）。方法US系统采用客户机／服务器（C／S）结构应用模式，硬件配有服务器1台，交换机1台，检验科工作站28台。应用MSSQL Server7．0数据库管理系统和Delphi5．0语言程序设计开发。US管理信息系统在应用上分为数据采集、样本管理、结果审核、检验仪器与信息资源共享和数据处理5个模块，同时具有打印检验报告单、进行质量控制、综合查询数据和数据安全性管理4个功能。结果运用US系统，实现了数据结果的实时自动接收、控制与综合分析，理顺了实验室的工作流程。与以往手工操作相比，申请方法更简捷、直观、高效，大大降低了误差率，比手工操作缩短3—4h，大大提高日工作效率，使医生能及时、准确看到检验结果数据。结论通过US系统的使用，提高了工作效率，减少误差，方便了医生临床诊断和患者就医，加强了医疗管理工作。     

家族性心肌病

家族性心肌病是一种亚急性或慢性的心肌病变。临床上主要可以分为肥厚型(非对称性心室间隔肥厚)、充血型和限制型三型。心律不齐和传导阻滞是本病的临床特征,而大多在出生后就可以发生。这种病人在40—50岁之间极易发生心力衰竭。本文报告一个家族性心肌病的家系。在该家系中,有一个值得注意的问题是非对称性心室间隔肥厚(肥厚型)常常与充血型同时存在。这种现象,Goodwin(1970)解释为肥厚型心肌病是充血型心肌病的早期阶段,也就是说充血型是肥厚型的发展结果。但从本家系的分析结果来看,充血型心肌病     

病房工作站联网运行发生差错的原因及防范措施

“军字 I号”工程医生、护士工作站在我院联网运行已1年 ,通过联机运行中经常出现的问题与差错 ,我们体会到不断强化微机操作培训 ,熟练掌握系统软件的功能与应用 ,严格执行网络程序及规章制度 ,加强医护配合与监督 ,落实检验与安全维护 ,是确保网络安全运行 ,提高病房工作站系统软件应用水平 ,防止发生医疗数据信息差错 ,杜绝医疗、护理差错事故发生的重要防范管理措施。1　病房工作站系统功能介绍1.1　医生工作站系统软件功能　医生工作站是面向临床医生 ,提供下达医嘱 ,书写病历 ,开申请单 ,查询报告单 ,查询体温单 ,填写首页 ,病历检索…     

眼科临床病案多媒体数据管理系统的设计及应用

结合眼科学疾病的临床诊断与标准,设计了基于Access的眼科学临床病案多媒体数据管理系统。对我院1998年1月–2011年12月眼科患者病案资料,进行了逻辑结构设计及功能模块设计。以Access2003构建眼科临床病案多媒体数据管理系统,实现了眼科各病案传统资料及图片、影音等多媒体资料的录入、存储、归类、查询、分析统计。该系统还实现了传统数据与多媒体数据的有效存储,功能强大,简单、实用,方便了病案的信息化管理,有利于临床诊疗、教学和科研。     

论建立人体数据库的必要性、可行性及实施方案

人体数据库是生命科学体系的一个不可分割的部分，电子计算机查询系统替代工具书是信息化的必然趋势，在广泛收集资料的基础上，对人体全方位数据按生理功能分类，利用数据库、多媒体光盘及网络技术制作成电子计算机查询系统软件，替代传统的工具书，使医疗及科研人员能够快速、准确地查找到所需最新数据。二十一世纪，人类将进入信息时代，同时生命科学将取代物理学而成为领导学科，人体数据库正是处于二者的交点上，是前进路上一块重要的基石，愿有识之士共同完成此综合性系统工程。     

An online locus-specific mutation database for familial hypertrophic cardiomyopathy.

The aim of this locus-specific mutation database was to provide an online resource that contains summarised and updated information on familial hypertrophic cardiomyopathy (FHC)-associated mutations and related data, for researchers and clinicians. It also serves as a means of publishing previously unpublished data, which could be of value in understanding genotype/phenotype correlations. There are 123 FHC-associated mutations catalogued along with ancillary information. By implementing the cgi/http method, remote users can query the database via the HTML interface on the Web browser and obtain data of relevance to them. The online service is available on http://www.angis.org.au/Databases/Heart.     

家族性肥厚型心肌病与肌钙蛋白T

家族性肥厚型心肌病(FHCM)是一种常染色体显性遗传病, 由于编码心肌蛋白的基因突变引起,目前已识别出至少13个不同致病基因的200余种突变。目前,阐明FHCM的分子遗传学机制已经成为当前研究的热点之一。肌钙蛋白T通过与原肌宁蛋白结合,在将肌钙蛋白复合体锚钉到细肌丝上起重要作用。肌钙蛋白T基因突变是导致家族性肥厚型心肌病的主要原因,至今已经发现了大约30个突变,约占所有突变的15~20%。肌钙蛋白T基因突变所致FHCM有两个主要特征：（1）心肌肥厚程度较轻, 疾病外显率差别较大,（2）猝死率高。目前所发现的致FHCM突变,主要集中在肌钙蛋白T的T1和T2结构域。对肌钙蛋白T突变致FHCM分子机制的研究将有助于肥厚型心肌病的基因诊断和临床治疗。     

EZ-Entry: a clinical data management system

In this paper, a new data management system named EZ-Entry is introduced. Five major functions are enclosed in this system: (1) user authentication; (2) database construction; (3) double data entry with instant alignment; (4) revision tracking; (5) query management. The practical application performed on two clinical trials indicates that EZ-Entry meets the requirements of clinical data management with high efficiency and security. This software is freely available on request from the authors for academic purposes.     

EURISWEB – Web-based epidemiological surveillance of antibiotic-resistant pneumococci in Day Care Centers

Background

EURIS (European Resistance Intervention Study) was launched as a multinational study in September of 2000 to identify the multitude of complex risk factors that contribute to the high carriage rate of drug resistant Streptococcus pneumoniae strains in children attending Day Care Centers in several European countries. Access to the very large number of data required the development of a web-based infrastructure – EURISWEB – that includes a relational online database, coupled with a query system for data retrieval, and allows integrative storage of demographic, clinical and molecular biology data generated in EURIS.

Methods

All components of the system were developed using open source programming tools: data storage management was supported by PostgreSQL, and the hypertext preprocessor to generate the web pages was implemented using PHP. The query system is based on a software agent running in the background specifically developed for EURIS.

Results

The website currently contains data related to 13,500 nasopharyngeal samples and over one million measures taken from 5,250 individual children, as well as over one thousand pre-made and user-made queries aggregated into several reports, approximately. It is presently in use by participating researchers from three countries (Iceland, Portugal and Sweden).

Conclusion

An operational model centered on a PHP engine builds the interface between the user and the database automatically, allowing an easy maintenance of the system. The query system is also sufficiently adaptable to allow the integration of several advanced data analysis procedures far more demanding than simple queries, eventually including artificial intelligence predictive models.     

心尖肥厚型心肌病患者心电图表现与临床诊断意义

目的：探讨心尖肥厚型心肌病患者的心电图特征性改变及临床诊断意义.方法：分析68例心尖肥厚型心肌病患者的心电图及超声心动图资料.结果：心尖肥厚心肌病合并有心电图异常和超声心动图异常改变者68例,中胸及左胸导联,巨大倒置酷似“冠状T波”≥0.2～0.3mV,以V3,V4明显且TV4＞TV5＞TV3者62例;左胸V4-V6导联R波电压增高,且RV4＞RV5＞RV6者60例;ST段压低以V3-V4最明显者62例;超声心动图示心尖肥厚达17mm或以上伴心尖部心腔狭小者68例.结论：心电图对心尖肥厚型心肌病具有早期诊断价值 。     

Mapping a gene for familial hypertrophic cardiomyopathy to chromosome 14q1

To identify the chromosomal location of a gene responsible for familial hypertrophic cardiomyopathy, we used clinical and molecular genetic techniques to evaluate the members of a large kindred. Twenty surviving and 24 deceased family members had hypertrophic cardiomyopathy; 58 surviving members were unaffected. Genetic-linkage analyses were performed with polymorphic DNA loci dispersed throughout the entire genome, to identify a locus that was inherited with hypertrophic cardiomyopathy in family members. The significance of the linkage detected between the disease locus and polymorphic loci was assessed by calculating a lod score (the logarithm of the probability of observing coinheritance of two loci, assuming that they are genetically linked, divided by the probability of detecting coinheritance if they are unlinked). A DNA locus (D14S26), previously mapped to chromosome 14 and of unknown function, was found to be coinherited with the disease in this family. No instances of recombination were observed between the locus for familial hypertrophic cardiomyopathy and D14S26, yielding a lod score of +9.37 (theta = 0). These data indicate that in this kindred, the odds are greater than 2,000,000,000:1 that the gene responsible for familial hypertrophic cardiomyopathy is located on chromosome 14 (band q1).     

Can sudden cardiac death be prevented?

Hypertrophic cardiomyopathy is regarded as the most common cause of sudden cardiac death in young people (including trained athletes). Introduction of implantable cardioverter-defibrillators to the hypertrophic cardiomyopathy patient population represents a new paradigm for clinical practice and perhaps the most significant advance in the management of this disease to date. Implantable defibrillators offer the only proven protection against sudden death by virtue of effectively terminating ventricular tachycardia/fibrillation and, in the process, altering the natural history of hypertrophic cardiomyopathy and providing the potential opportunity of normal or near-normal longevity for many patients. However, targeting the most appropriate candidates for prophylactic device therapy can be complex, compounded by the unpredictability of the underlying arrhythmogenic substrate, absence of a single dominant and quantitative risk marker in this heterogeneous disease, and the historical difficulty in assembling sufficiently powered prospective and randomized trials in large patient populations. Nevertheless, the current risk factor algorithm, when combined with a measure of individual physician judgment, is an effective strategy for identifying high-risk patients. Indeed, prevention of sudden death has now become an integral, albeit challenging, component of overall hypertrophic cardiomyopathy management.     

Determining pathogenicity of genetic variants in hypertrophic cardiomyopathy: importance of periodic reassessment

PurposeMajor advances have been made in our understanding and clinical application of genetic testing in hypertrophic cardiomyopathy. Determining pathogenicity of a single-nucleotide variant remains a major clinical challenge. This study sought to reassess single-nucleotide variant classification in hypertrophic cardiomyopathy probands.MethodsConsecutive probands with hypertrophic cardiomyopathy with a reported pathogenic mutation or variation of uncertain significance were included. Family and medical history were obtained. Each single-nucleotide variant was reassessed by a panel of four reviewers for pathogenicity based on established criteria together with updated cosegregation data and current population-based allele frequencies.ResultsFrom 2000 to 2012, a total of 136 unrelated hypertrophic cardiomyopathy probands had genetic testing, of which 63 (46%) carried at least one pathogenic mutation. MYBPC3 (n = 34; 47%) and MYH7 (n = 23; 32%) gene variants together accounted for 79%. Five variants in six probands (10%) were reclassified: two variation of uncertain significance were upgraded to pathogenic, one variation of uncertain significance and one pathogenic variant were downgraded to benign, and one pathogenic variant (found in two families) was downgraded to variation of uncertain significance. None of the reclassifications had any adverse clinical consequences.ConclusionGiven the rapid growth of genetic information available in both disease and normal populations, periodic reassessment of single-nucleotide variant data is essential in hypertrophic cardiomyopathy.Genet Med 2014:16(4):286–293.     

Clinical predictors of genetic testing outcomes in hypertrophic cardiomyopathy

PurposeGenetic testing for hypertrophic cardiomyopathy has been commercially available for almost a decade; however, low mutation detection rate and cost have hindered uptake. This study sought to identify clinical variables that can predict probands with hypertrophic cardiomyopathy in whom a pathogenic mutation will be identified.MethodsProbands attending specialized cardiac genetic clinics across Australia over a 10-year period (2002–2011), who met clinical diagnostic criteria for hypertrophic cardiomyopathy and who underwent genetic testing for hypertrophic cardiomyopathy were included. Clinical, family history, and genotype information were collected.ResultsA total of 265 unrelated individuals with hypertrophic cardiomyopathy were included, with 138 (52%) having at least one mutation identified. The mutation detection rate was significantly higher in the probands with hypertrophic cardiomyopathy with an established family history of disease (72 vs. 29%, P < 0.0001), and a positive family history of sudden cardiac death further increased the detection rate (89 vs. 59%, P < 0.0001). Multivariate analysis identified female gender, increased left-ventricular wall thickness, family history of hypertrophic cardiomyopathy, and family history of sudden cardiac death as being associated with greatest chance of identifying a gene mutation. Multiple mutation carriers (n = 16, 6%) were more likely to have suffered an out-of-hospital cardiac arrest or sudden cardiac death (31 vs. 7%, P = 0.012).ConclusionFamily history is a key clinical predictor of a positive genetic diagnosis and has direct clinical relevance, particularly in the pretest genetic counseling setting.Genet Med15 12, 972–977.