电子束计算机断层摄影术在小儿心脏病诊断中的应用

目的评价电子束计算机断层摄影术(EBCT)在小儿心脏病诊断中的价值.方法对78例患儿经胸常规超声心动图(TTE)诊断为小儿心脏病,其中74例为先天性心脏病,复杂先天性心脏病占63例,非先天性心脏病变4例,同时做EBCT增强单层容积扫描,并由计算机工作站进行图像三维重建.其中25例行常规心血管造影,43例经外科手术,进行对照研究.结果全组20种小儿先天性心脏病共176处畸形,其中复杂先天性心脏病16种.心内畸形共65处,TTE漏诊1处(三尖瓣闭锁),TTE诊断准确率98.5%.而心外大血管畸形及心室-大血管连接异常诊断中两者有明显差别,EBCT诊断准确率为96.4%(107/111),TTE诊断准确率86.5%(96/111).EBCT与TTE诊断的准确率差异非常显著(χ2=6.964,P＜0.01).手术纠治心脏病变共109处,EBCT确诊105处,诊断准确率96.3%(105/109).TTE确诊97处,诊断准确率89.0%(97/109).EBCT与TTE诊断的准确率有显著差异(χ2=4.318,P＜0.05).TTE、EBCT对4例非先天性心脏病均作出正确诊断.TTE与EBCT结合使诊断的总准确率提高到99.4%(175/176).结论E-BCT对小儿心脏病变的检出优于TTE.EBCT血管造影与TTE及心血管造影相结合,可提高对小儿心脏病诊断的准确率,以指导手术或药物治疗.     

利用神经网络构建超声多普勒栓子信号自动检测系统

对血液中栓子的早期检测有着重要的临床诊断意义.超声多普勒技术使栓子的无损检测成为可能,但是目前的检测尚依赖于医生的手动操作和临床经验,仍缺乏可靠的自动检测系统.本文基于小波包变换和主元分析提取对栓子敏感的特征参数,并利用神经网络构建超声多普勒栓子信号的自动检测系统.仿真研究和临床实验表明:该系统有效提高了栓子检测的准确性,有望应用于临床诊断.     

改进U-Net深度网络的视网膜血管分割算法

为了解决U-Net算法在分割眼底图像时无法分割末梢微小血管和无法处理噪声干扰等问题,提出了一种改进的视网膜血管分割算法。首先,在U-Net算法中引入通道强化残差网络,用以优化U-Net架构,使得网络识别更多视网膜微血管。其次,引入空间注意力网络来排除噪声,更好地突出血管。最后,在损失函数的计算中,使用动态权重代替U-Net算法的固定权重,迫使神经网络能够学习一个稳健的特征映射。将改进的算法在DRIVE数据集上进行实验,实验结果表明本文分割算法的准确性和敏感性大幅提高。比原U-Net算法准确性和敏感性分别提高了2.12%和7.51%,比DCU-Net准确性和敏感性分别提高了1.20%和2.55%。     

超声弹性成像联合常规超声在甲状腺结节良恶性鉴别中的诊断价值探究

目的:分析超声弹性成像(Ultrasound Elastography,UE)联合常规超声在甲状腺结节良恶性鉴别诊断中的应用价值.方法:选取本院 60 例甲状腺结节患者作为研究对象,均行常规超声与UE检查,以病理结果为诊断的"金标准",比较常规超声与UE检查单独与联合鉴别恶性甲状腺结节的敏感度、特异度及准确性.结果:常规超声鉴别恶性甲状腺结节的敏感度为 68.75%,特异度为 77.27%,准确性为 75.00%;UE鉴别敏感度为 62.50%,特异度为 81.82%,准确性为76.67%;UE联合常规超声鉴别敏感度 93.75%,特异度 95.45%,准确性 95.00%;常规超声联合UE诊断特异度、准确性明显高于常规超声、UE单一诊断(P＜0.05),诊断敏感度明显高于UE单独应用(P＜0.05).结论:UE联合常规超声可有效提高甲状腺结节良恶性鉴别诊断特异度与准确性.     

心脏B超和心电图诊断高血压性心脏病的准确性对比

目的探讨高血压性心脏病采用心脏B超和心电图诊断的临床准确性。方法选取我院高血压性心脏病患者100例,分为两组。对照组50例作心电图检测;观察组50例作心脏B超检测。对比分析检查的准确率。结果观察组检测的准确率88.0%(44/50)明显高于对照组的58.0%(29/50),差异具有统计学意义(P0.01);观察组所检出的44例中,对称性肥厚16例(32.0%),非对称性室间隔肥厚28例(56.0%)。结论高血压性心脏病实施心脏B超检测,准确率高,能真实反映解剖结构,诊断价值较高,值得推广应用。     

单纯性先天性心脏病致病基因的研究进展

先天性心脏病(congenital heart disease,CHD)是胎儿时期心血管发育异常而导致的先天性畸形疾病,在新生儿非感染性死亡中占首位.其中约有80%仅表现为心脏畸形,而不伴有其他系统的先天异常,称之为单纯性先天性心脏病.近年来,用分子遗传学的手段来研究单纯性先天性心脏病的分子基础已经取得了较大进展,从而促进了其临床诊断及治疗方法的拓展.     

基于栈式深度多项式网络集成学习框架的帕金森病计算机辅助诊断

特征表达是基于磁共振成像(MRI)的帕金森病(PD)计算机辅助诊断系统诊断准确性的重要决定因素。深度多项式网络(DPN)是一种新的有监督深度学习算法,对于小数据集具有良好的特征表达能力。本文提出一种面向PD计算机辅助诊断的栈式DPN(SDPN)集成学习框架,以有效提高基于小数据的PD辅助诊断准确性。本框架对所提取的MRI特征的每一个特征子集分别通过SDPN得到新的特征表达,然后采用支持向量机(SVM)对每个子集进行分类,再对所有分类器进行集成学习,得到最终的PD诊断结果。通过对公开的帕金森病数据库PPMI进行实验,基于脑网络特征的分类精度、敏感度和特异性分别为90.15%、85.48%和93.27%;而基于多视图脑区特征的分类精度、敏感度和特异性分别为87.18%、86.90%和87.27%。与在PPMI数据库中的MRI数据集进行实验的其他算法研究相比,本文所提出的算法获得了更好的分类结果。本文研究表明了所提出的SDPN集成学习框架的有效性,具有应用于PD计算机辅助诊断的可行性。     

基于轻量级神经网络的新冠肺炎CT新型识别技术

目的:为了满足临床新冠肺炎检测的实际需求,提出一种基于轻量级人工神经网络的新冠肺炎CT新型识别技术。方法:首先,选取目前公开的所有新冠肺炎CT图像数据集,经过图像亮度规范化和数据集清洗后作为训练数据,通过大样本提高深度学习的泛化能力;其次,采用GhostNet轻量级网络简化网络参数,使深度学习模型能够在医用计算机上运行,提高新冠肺炎CT诊断的效率;再次,在网络输入中加入肺部区域分割图像,进一步提高新冠肺炎CT诊断的准确性;最后,提出加权交叉熵损失函数减小漏诊率。结果:在本研究构建的数据集上进行测试,所提出方法的精确率、召回率、准确率和F1值分别为83%、96%、90%和88%,且在医用计算机上耗时为236 ms。结论:本研究提出方法的效率和准确性均优于其他对比算法,能较好地适应新冠肺炎诊断的需求。     

血清β-CTX和BGP联合应用在乳腺癌骨转移诊断中的意义

目的:探讨血清Ⅰ型胶原羧基端肽β特殊序列(β-CTX)和骨钙素(BGP)定量检测诊断乳腺癌骨转移的价值。方法:用电化学发光免疫分析(ECLIA)测定35例乳腺癌骨转移患者及30例乳腺癌无骨转移患者血清β-CTX和BGP水平。结果:乳腺癌骨转移组患者血清β-CTX和BGP水平均明显高于无骨转移组和对照组(P<0.01),无骨转移组与对照组间无显著差异(P>0.05)。血清β-CTX诊断乳腺癌骨转移的敏感性、特异性和准确性分别为62.3%、88.5%、82.7%;血清BGP诊断乳腺癌骨转移的敏感性、特异性和准确性分别为75.6%、92.1%、83.5%。两者联合诊断乳腺癌骨转移的敏感性、特异性和准确性分别为91.8%、87.6%、84.2%。结论:血清β-CTX和BGP对乳腺癌骨转移的诊断均有重要价值,两者联合有助于提高乳腺癌骨转移的敏感性和准确性。     

一种基于足压数据主成分分析的足弓形态智能检测方法

目的 根据临床中平足与高弓足量化评估的要求,提出一种基于足压数据主成分分析（principal component analysis, PCA）智能快速足弓形态检测方法,并验证其临床有效性。方法 纳入诊断为足弓异常与足弓健康的志愿者,设计研发一套便携式足弓智能检测系统。采用44×52阵列式薄膜压阻传感器,采集静态站立式足底压力分布数据,利用自行编写的PCA算法自动拟合足轴线,进行足弓诊断并生成诊断报告。将足压采集结果与现有设备进行比对,验证足压数据的准确性。对于平足、高弓足和正常3类足弓的判别算法,通过对比临床诊断验证评估准确性。结果 该系统与现有压力采集设备的测量结果具有较好的相关性,接触面积偏差低于3.2%,计算拟合的足轴线与临床定义角度偏差小于1°,且该系统能获得与临床中足弓形态诊断相符率92.6%的评估结果。结论 引入PCA对足轴线自动化拟合,实现了快速而准确提取足弓信息的目的。该方法可用于临床实践中平足与高弓足的辅助筛查,有助于开展足弓畸形程度的量化分析和病理机制的研究。     

先天性心脏病筛查系统设计

为提高医生筛查先天性心脏病的效率,设计一款基于卷积神经网络的先天性心脏病筛查系统。系统以软硬协同的方式实现心音、心电等生理参数的实时同步采集以及可视化和定量化分析。系统包含上下位机,下位机以FPGA为核心实现心音心电数据采集以及小波阈值去噪等预处理,上位机在Windows系统环境下以Python编程语言实现二阶谱特征提取、卷积神经网络二分类识别以及用户界面可视化显示。最终,系统对200名志愿者进行测试,准确率达到94.5%,特异度为95.9%,敏感度为93.2%。结果表明系统具有良好的表现,可以为临床先心病筛查提供有效的辅助。     

基于机器学习的心音识别分类研究

心音信号可反映心脏的病理信息,是诊断心脏健康的重要依据之一。本文首先从心音信号提取时频域、梅尔倒谱系数等145个特征作为机器学习的输入数据集,然后在随机森林、LightGBM、XGBoost、GBDT、SVM共5种分类器中选出效果最佳分类器与递归特征消除算法结合进行数据挖掘,找出重要特征集并对其分类效果做比较与分析,最后运用Stacking模型融合方法优化模型。数据挖掘特征子集比同数量特征子集在准确率、召回率、精确率、F1值上分别提高了33.51%、14.54%、20.61%、24.04%;采用LightGBM和SVM模型融合可将F1值提高至92.6%。本文提出了一种有效的心音识别分类方法,挖掘出心音最重要的8个特征,为临床诊断提供参考。     

A comparison of different feature extraction methods for diagnosis of valvular heart diseases using PCG signals

This article presents a novel method for diagnosis of valvular heart disease (VHD) based on phonocardiography (PCG) signals. Application of the pattern classification and feature selection and reduction methods in analysing normal and pathological heart sound was investigated. After signal preprocessing using independent component analysis (ICA), 32 features are extracted. Those include carefully selected linear and nonlinear time domain, wavelet and entropy features. By examining different feature selection and feature reduction methods such as principal component analysis (PCA), genetic algorithms (GA), genetic programming (GP) and generalized discriminant analysis (GDA), the four most informative features are extracted. Furthermore, support vector machines (SVM) and neural network classifiers are compared for diagnosis of pathological heart sounds. Three valvular heart diseases are considered: aortic stenosis (AS), mitral stenosis (MS) and mitral regurgitation (MR). An overall accuracy of 99.47% was achieved by proposed algorithm.     

高血压性心脏病心脏B超与心电图的对比观察

目的临床观察对比高血压性心脏病采用心脏B超与心电图检查的效果。方法选取48例经诊断为高血压性心脏病患者分别采用心电图检查诊断(心电组)与超声超声仪器进行诊断(超声组),观察两组诊断准确度。结果超声组左室扩大有18例占37.5%,左室肥厚37例占77.1%,左房增大有19例占39.6%;心电组左室扩大有10例占20.8%,左室肥厚17例占35.4%,左房增大有6例占12.5%。高血压性心脏病检测准确度超声组较心电组高,P0.05。结论临床对高血压性心脏病患者采用心脏B超检查不仅可清楚了解患者心脏情况,且诊断准确度高。     

The impact of technology on diagnostic radiology

Summary

In conventional radiology the introduction of image intensifiers and television has contributed greatly to the accuracy of radiological diagnosis in a number of fields. For example in diseases of the heart and vascular system the development of sophisticated radiological methods of diagnosis has assisted in the spectacular advances in cardio-vascular surgery.

The introduction of non-invasive methods of imaging including CT scanning, diagnostic ultrasound, scintiscanning and nuclear magnetic resonance has substantially assisted in patient management.

The impact of CT scanning in diseases of the nervous system has been spectacular and it is likely that scanning by nuclear magnetic resonance will make a similar contribution. Diagnostic ultrasound has virtually replaced conventional radiography in obstetrics and its value is being established in other systems. Investigation by radioisotopes is very useful particularly in deseases affecting the lungs such as pulmonary infarction. Investigations using radioisotopes are of importance in assessing functional abnormalities of the lungs, heart, kidneys and in patients where the flow of cerebrospinal fluid is impaired.     

基于极限梯度提升和深度神经网络共同决策的心音分类方法

心音是诊断心血管疾病常用的医学信号之一。本文对心音正常/异常的二分类问题进行了研究,提出了一种基于极限梯度提升(XGBoost)和深度神经网络共同决策的心音分类算法,实现了对特征的选择和模型准确率的进一步提升。首先,本文对预处理后的心音信号进行心音分割,在此基础上提取了5个大类的特征,前4类特征采用递归特征消除法进行特征选择,作为XGBoost分类器的输入,最后一类为梅尔频率倒谱系数(MFCC),作为长短时记忆网络(LSTM)的输入。考虑到数据集的不平衡性,本文在两种分类器中皆使用了加权改进的方法。最后采用异质集成决策方法得到预测结果。将本文所提心音分类算法应用于PhysioNet网站在2016年发起的PhysioNet心脏病学挑战赛(CINC)所用公开心音数据库,以测试灵敏度、特异性、修正后的准确率以及F得分,结果分别为93%、89.4%、91.2%、91.3%,通过与其他研究者应用机器学习、卷积神经网络(CNN)等方法的结果比较,在准确率和灵敏度上有明显提高,证明了本文方法能有效地提高心音信号分类的准确性,在部分心血管疾病的临床辅助诊断应用中有很大的潜力。     

新发传染病鉴别诊断及临床诊断辅助识别系统的初步应用

目的 检验所建立的新发传染病鉴别诊断及临床诊断辅助识别系统诊断提示功能与临床的符合性.方法 应用所建立的新发传染病鉴别诊断及临床诊断辅助识别系统对49种314份感染性疾病病例进行回顾性研究,对186例初诊发热患者进行前赡性研究.结果 回顾性对验证结果 显示,系统第一诊断提示与临床符合率61.9%;前3位提示符合率78.1%;前5位提示符合率86.6%.初诊发热患者进行前瞻性研究,系统第1诊断提示与临床符合率59.7%;前2位提示符合率77.9%;前5位提示符合率85.4%.结论 该系统对新发传染病(EID)诊断与鉴别诊断有很好的提示作用,具有很好的实用价值.     

An update on Kawasaki disease

Kawasaki disease (KD) is a febrile systemic vasculitis complicated by coronary and peripheral arterial aneurysms in 20-35% of untreated patients. It is reported as the commonest cause of acquired heart disease in children in developed countries, and may be a risk for adult ischaemic heart disease. Although KD has been reported all over the world, it is overexpressed among Asian populations, especially Japanese. The disease pathogenesis is still unknown and several theories have been proposed, including the possibility of an infection by a toxin-secreting microorganism and of a superantigen-driven process. Despite numerous efforts there is still no diagnostic test available for KD, and the diagnosis is based on clinical criteria after the exclusion of other diseases presenting with high persistent fever. Prompt diagnosis is critical, since the early administration of intravenous immunoglobulins and aspirin reduces the rate of coronary abnormalities to less than 5% of patients.     

An intelligent model for liver disease diagnosis

Objectives: Liver disease, the most common disease in Taiwan, is not easily discovered in its initial stage; early diagnosis of this leading cause of mortality is therefore highly important. The design of an effective diagnosis model is therefore an important issue in liver disease treatment. This study accordingly employs classification and regression tree (CART) and case-based reasoning (CBR) techniques to structure an intelligent diagnosis model aiming to provide a comprehensive analytic framework to raise the accuracy of liver disease diagnosis. Methods: Based on the advice and assistance of doctors and medical specialists of liver conditions, 510 outpatient visitors using ICD-9 (International Classification of Diseases, 9th Revision) codes at a medical center in Taiwan from 2005 to 2006 were selected as the cases in the data set for liver disease diagnosis. Data on 340 patients was utilized for the development of the model and on 170 patients utilized to perform comparative analysis of the models. This paper accordingly suggests an intelligent model for the diagnosis of liver diseases which integrates CART and CBR. The major steps in applying the model include: (1) adopting CART to diagnose whether a patient suffers from liver disease; (2) for patients diagnosed with liver disease in the first step, employing CBR to diagnose the types of liver diseases. Results: In the first phase, CART is used to extract rules from health examination data to show whether the patient suffers from liver disease. The results indicate that the CART rate of accuracy is 92.94%. In the second phase, CBR is developed to diagnose the type of liver disease, and the new case triggers the CBR system to retrieve the most similar case from the case base in order to support the treatment of liver disease. The new case is supported by a similarity ratio, and the CBR diagnostic accuracy rate is 90.00%. Actual implementation shows that the intelligent diagnosis model is capable of integrating CART and CBR techniques to examine liver diseases with considerable accuracy. The model can be used as a supporting system in making decisions regarding liver disease diagnosis and treatment. The rules extracted from CART are helpful to physicians in diagnosing liver diseases. CBR can retrieve the most similar case from the case base in order to solve a new liver disease problem and can be of great assistance to physicians in identifying the type of liver disease, reducing diagnostic errors and improving the quality and effectiveness of medical treatment.