基于交叉对比神经网络的心音分类

目的:通过交叉对比神经网络（CCNN）实现心音信号的自动分类，从而对心血管疾病进行早期诊断。方法:实验基于PhysioNet/Cinc 2016心音数据库。训练集和测试集数据来自互斥的健康受试者/病理患者，并以4:1的比例进行划分，输入CCNN。CCNN利用深度卷积神经网络进行特征提取，结合基于信息的相似度度量理论（IBS），对特征向量间的相似性进行度量并分类。结果:实验结果得出灵敏度为0.834 6，特异性为0.962 3，最终大赛综合得分为0.898 5。结论:CCNN使用交叉对比的输入模式扩充数据量，引入信号间的对比信息，同时在神经网络的训练过程中应用统计学思想，使网络具备良好的泛化性，更加适应医学数据量较少的场景，在心音分类中取得较好的结果。     

Unexpected death in a young child associated with anomalous aortic origin of the left main coronary artery without physical exertion: A case of an anomalous coronary artery with highly abundant elastic fibers

Sudden death due to anomalous aortic origin of a coronary artery is far less common among young children in the absence of exercise stress. This report describes the case of a 2-year-old boy with a lower respiratory tract infection who suffered sudden cardiac arrest in his bed at home. The autopsy revealed that the left coronary artery (LCA) originated from the right sinus of Valsalva with an acute angle takeoff and traveled between the aorta and the pulmonary trunk (an interarterial course). Upon histological examination, the LCA, before reaching its major branches, was located adjacent to the outside of the aortic wall without an intramural passage, and the arterial wall was composed almost exclusively of elastic fibers without media containing smooth muscle cells throughout the entire length of the abnormal running. Screening tests for respiratory virus infection detected enterovirus in the lung tissue. In association with an acute angle takeoff and interarterial course, the wall structure with highly abundant elastic fibers that are more flexible tissues among blood vessel components might suggest their vulnerability to compression during the great vessels’ systolic expansion in the sympathetic activation induced by the viral infection, leading to fatal myocardial ischemia without physical exertion.     

Cerebellar artery infarction with sudden hearing loss and vertigo as initial symptoms: A case report

BACKGROUNDSudden hearing loss (SHL) is associated with serious systematic conditions such as neoplasms, vascular events, autoimmune diseases, infections, and iatrogenic injury. Some authors report that SHL can be an early warning sign of impending vertebrobasilar ischemic stroke. It is important to distinguish stroke from benign disease. CASE SUMMARYA 48-year-old male patient presented with SHL and vertigo as first symptoms. Diffusion-weighted imaging revealed high signal intensity in the left posterior inferior cerebellar artery territory of the cerebellar hemisphere and high signal intensity in the right pons and bridge cerebellar arm, confirming that the patient had cerebral infarction. Treatment with antiplatelet drugs, steroid anti-inflammatory drugs, and neurotrophic nerve therapy promoted blood circulation and removed blood stasis, and the symptoms of the patient were significantly improved. CONCLUSIONSHL and vertigo could be the initial symptoms of vertebrobasilar ischemic stroke.     

关节镜下经髌腱入路空心拉力螺钉复位固定治疗胫骨髁间嵴撕脱骨折*

目的探讨在关节镜下经髌腱入路,利用空心拉力螺钉复位固定胫骨髁间嵴撕脱骨折(TEFx)的安全性和有效性。方法选取2014年1月-2015年12月23例TEFx的患者,均在关节镜下经髌腱入路,用空心拉力螺钉复位固定。术前Meyers-McKeever分型:Ⅱ型8例,Ⅲ型10例,Ⅳ型5例;男17例,女6例;年龄16～53岁,平均27.8岁。术前前抽屉试验、Lachman试验均阳性。比较术前术后的视觉模拟评分(VAS)、Lysholm、Tegner和国际膝关节文献委员会(IKDC)评分评价患侧膝关节功能。结果 23例患者均得到随访,随访时间30～40个月,平均36个月。术后即刻X线片示TEFx均复位良好,术后3个月骨折均愈合。无1例感染、关节僵硬、伸直受限、复位丢失及神经血管损伤等并发症。最终随访患侧膝关节活动度均恢复正常,前抽屉试验、Lachman试验均阴性。VAS评分术前(4.8±1.2)分,最终随访为(1.2±0.8)分,术前术后比较,差异有统计学意义(t=18.72,P=0.003);Lysholm评分术前为(50.8±6.2)分,最终随访为(90.8±5.4)分,术前术后比较,差异有统计学意义(t=-42.64,P=0.000);Tegner评分术前为(4.0±1.0)分,最终随访为(5.1±1.2)分,术前术后比较,差异有统计学意义(t=-16.82,P=0.005);IKDC主观评分术前为(52.5±7.4)分,最终随访为(91.5±5.7)分,术前术后比较,差异有统计学意义(t=-40.58,P=0.000)。结论膝关节镜下经髌腱入路空心拉力螺钉内固定治疗TEFx具有微创、操作简捷、固定可靠和恢复快的优点。     

Assessing the utility of yearly pre-season laboratory screening for athletes on a major professional sports team

Objectives

Professional athletes undergo annual pre-season laboratory screening, although clinical evidence supporting the practice is limited and no uniform set of guidelines on pre-season laboratory screening exists. The aim of this study was to assess the clinical value of annual pre-season laboratory screening tests for a major professional sports team over multiple years.

DVH预测模型在VMAT计划培训中的应用

目的:以宫颈癌为例，探索使用剂量体积直方图（DVH）预测模型培训放疗物理师容积旋转调强（VMAT）计划设计的可行性及其效果。方法:随机选取20例宫颈癌测试病例对3名见习物理师进行两个阶段培训（第一阶段15例，第二阶段5例）。每位物理师分别采用两种方法设计VMAT计划:传统人工计划（MP）和基于DVH预测模型引导的自动计划（KBP）。对比不同阶段两种计划靶区和危及器官的各种剂量学参数，并做相应统计分析。结果:与人工计划相比，自动计划在明显提高PGTV靶区V60覆盖（P<0.001），改善高剂量控制（V66）（P=0.027）的情况下，显著降低膀胱（P<0.001）、直肠（P<0.001）、左右肾（P=0.001和P<0.001）以及左右侧股骨头（P<0.001和P<0.001）等绝大部分正常组织器官的受照剂量。在提高计划合格率的同时，亦减小了计划者间的质量差异。结论:DVH预测模型有助于深化物理师对VMAT初始优化参数设置的理解，加快培训进度，提升VMAT计划设计水平。     

Role of subcutaneous implantable loop recorder for the diagnosis of arrhythmias in Brugada syndrome: A United Kingdom single-center experience

1. Download : Download high-res image (285KB)
2. Download : Download full-size image

     

基于Bark域的电子耳蜗频带划分分析和拟合研究

目前电子耳蜗主流的言语处理策略是基于滤波器组的言语处理算法，该算法通过分频带进行信号处理并把参数传递到对应的电极上。电子耳蜗滤波器组的频带划分不是等分的，而是按一定规律进行并且符合人耳听觉特性的，其中，Bark域的频带划分是重要的参考。本研究基于Bark域的电子耳蜗频带划分方法，探讨Bark域频带划分的特性并结合目前的电子耳蜗滤波器组的频带划分进行分析，进而探讨频带划分中的曲线拟合方法，为电子耳蜗滤波器组中的频带划分提供重要的方法和参数。