基于预测可行性DVH的肝癌VMAT自动计划方法研究

目的 建立基于预测可行性剂量体积直方图(DVH)的肝癌容积调强弧形治疗(VMAT)自动计划方法,并评价其性能。方法 回顾性随机选取10例放疗肝癌病例。采用Pinnacle Auto-Planning设计VMAT自动计划,通过PlanIQ剂量预测得到可行性DVH曲线,并根据其显示的可行性目标区间设置自动计划的初始优化目标。评价计划靶体积、正常肝和其他危及器官的剂量参数以及机器跳数,并与临床手工计划行配对t检验。结果 自动计划和手工计划的计划靶体积D2%、D98%、Dmean和均匀度指数相近[(58.55±2.81) Gy∶(57.98±4.17) Gy、(47.15±1.58) Gy∶(47.82±1.38) Gy、(53.14±0.95) Gy∶(53.44±1.67) Gy和1.15±0.05∶1.14±0.07,P均>0.05],手工计划的计划靶体积适形指数略高于自动计划(0.77±0.08∶0.69±0.06,P<0.05)。自动计划的肝V30Gy、V20Gy、V10Gy、V5Gy和V<5Gy显著优于手工计划[(26.68±11.13)%∶(28.00±10.95)%、(29.96±11.50)%∶(31.89±11.51)%、(34.88±11.51)%∶(38.66±11.67)%、(45.38±12.40)%∶(50.74±13.56)%和(628.52±191.80) cm³∶(563.15±188.39) cm³,P均<0.05],自动计划的小肠、十二指肠、心脏Dmean以及全肺V10Gy低于手工计划[(1.83±2.17) Gy∶(2.37±2.81) Gy、(9.15±9.36) Gy:(11.18±10.49) Gy、(5.44±3.10) Gy∶(6.25±3.26) Gy以及(12.70±7.08)%∶(14.47±8.11)%,P均<0.05]。两种计划的机器跳数相近[(710.67±163.72) MU∶(707.53±155.89) MU,P>0.05]。结论 基于预测可行性DVH的肝癌VMAT自动计划方法能提高计划质量,更好保护正常肝,对小肠、十二指肠、全肺和心脏的保护也有优势。     

基于CT增强扫描图像的灰度直方图纹理分析评估胰腺癌分化程度

目的探究CT增强图像灰度直方图纹理分析技术在术前评估胰腺癌恶性程度的可行性，评价灰度直方图参数和胰腺癌病理分级、分化程度的关系。 方法回顾性分析扬州大学附属苏北人民医院术后病理证实的胰腺癌患者49例，其中高分化组27例，中-低分化组22例。选取2组肿瘤CT增强扫描静脉期横断位图像最大层面，采用Mazda软件勾画感兴趣区，并进行灰度直方图纹理分析，对高分化、中-低分化两组直方图特征参数进行统计分析。对有统计学意义的参数建立受试者工作特征曲线(receiver operating characterist，ROC)，最终对各参数其诊断效能进行比较分析。此外，对差异有统计学意义的参数与高分化、中-低分化组别之间的相关性进行检测。 结果采用灰度直方图分析提取的9个参数特征中，其中均值(Mean)、第1百分位数(Perc.01%)、第10百分位数(Perc.10%)、第50百分位数(Perc.50%)在高分化、中-低分化2组的差异有统计学意义(P<0.05)，变异度(Variance)、偏度(Skewness)、峰度(Kurtosis)、第90百分位数(Perc.90%)、第99百分位数(Perc.99%)在2组中差异无统计学意义(P>0.05)。采用Spearman相关性分析可知均值、Perc.01%、Perc.10%、Perc.50%与分化程度之间呈正相关(r值分别为0.04、0.29、0.32、0.33，P均<0.05)。评估胰腺癌恶性程度时，均值的诊断效能最高(AUC＝0.695)；当均值取值147.27时，其对应的敏感度和特异度分别为86.4%、44.4%；当Perc.01%取值117时，其对应的敏感度和特异度分别为95.5%、40.7%；当Perc.10%取值120时，其对应的敏感度和特异度分别为72.7%、63%；当Perc.50%取值146时，其对应的敏感度和特异度分别为86.4%、44.4%。 结论CT增强图像灰度直方图分析的特征参数在高分化、中-低分化胰腺癌之间存在差异，给术前评估胰腺癌恶性程度提供了新的方法。     

Deciphering the contribution of intrinsic and synaptic currents to the effects of transient synaptic inputs on human motor unit discharge

The amplitude and time course of synaptic potentials in human motoneurons can be estimated in tonically discharging motor units by measuring stimulus-evoked changes in the rate and probability of motor unit action potentials. However, in spite of the fact that some of these techniques have been used for over 30 years, there is still no consensus on the best way to estimate the characteristics of synaptic potentials or on the accuracy of these estimates. In this review, we compare different techniques for estimating synaptic potentials from human motor unit discharge and also discuss relevant animal models in which estimated synaptic potentials can be compared to those directly measured from intracellular recordings. We also review the experimental evidence on how synaptic noise and intrinsic motoneuron properties influence their responses to synaptic inputs. Finally, we consider to what extent recordings of single motor unit discharge in humans can be used to distinguish the contribution of changes in synaptic inputs versus changes in intrinsic motoneuron properties to altered motoneuron responses following CNS injury.     

A contrast correction method for dental images based on histogram registration

Contrast correction is often required in digital subtraction radiography when comparing medical data acquired over different time periods owing to dissimilarities in the acquisition process. This paper focuses on dental radiographs and introduces a novel approach for correcting the contrast in dental image pairs.The proposed method modifies the subject images by applying typical registration techniques on their histograms. The proposed histogram registration method reshapes the histograms of the two subject images in such a way that these images are matched in terms of their contrast deviation. The method was extensively tested over 4 sets of dental images, consisting of 72 registered dental image pairs with unknown contrast differences as well as 20 dental pairs with known contrast differences. The proposed method was directly compared against the well-known histogram-based contrast correction method.The two methods were qualitatively and quantitatively evaluated for all 92 available dental image pairs. The two methods were compared in terms of the contrast root mean square difference between the reference image and the corrected image in each case. The obtained results were also verified statistically using appropriate t-tests in each set.The proposed method exhibited superior performance compared with the well-established method, in terms of the contrast root mean square difference between the reference and the corrected images. After suitable statistical analysis, it was deduced that the performance advantage of the proposed approach was statistically significant.     

基于增强CT全瘤直方图分析鉴别诊断≤5 cm胃肠道间质瘤与胃神经鞘瘤

目的 探讨基于增强CT全瘤直方图分析鉴别诊断最大径≤5 cm胃肠道间质瘤（GIST）与胃神经鞘瘤（GS）的价值。方法 回顾性分析在行腹部增强CT检查并经手术病理证实的最大径≤5 cm GIST患者51例和GS患者16例。以MaZda软件在包含肿瘤的每一层图像上手动勾画瘤体ROI，进行直方图分析，获得动脉期和静脉期图像的均值、变异度、偏度、峰度及第1、10、50、90、99百分位数9个参数。比较2组各参数的差异并分析参数的诊断效能。结果 动脉期直方图参数中≤5 cm GIST的均值、变异度和第1、10、50、90百分位数分别为125.71±13.57、145.31（110.44，191.84）、97.63±16.40、110.45±14.69、125.67±13.46、141.06±13.48，GS的各值分别为139.97±19.08、99.21（80.97，135.21）、113.38±19.86、127.44±18.90、140.25±19.20、152.75±20.00，2组间差异均有统计学意义（P均<0.05），ROC曲线的AUC分别为0.76、0.72、0.75、0.78、0.76和0.73。静脉期直方图参数中≤5 cm GIST变异度为144.08（106.58，221.25），GS为94.38（80.97，122.90），前者大于后者（P<0.05），AUC为0.78。结论 增强CT直方图对鉴别≤5 cm GIST与GS有较高价值，其中动脉期第10%百分位数和静脉期变异度的诊断效能较高。     

基于多颜色空间和场均直方图的视频场景分类

视频分析中的一个热点和难点是研究视频的语义,例如场景分类、视频检索、移动目标行为分析等。其中场景分类是视频语义研究中的一个基础问题。在对视频场景进行分割的基础上,通过分析场景多个颜色分量的统计直方图,发现大多数典型场景总有一个或者多个颜色分量的场均直方图具有独特性,即在某个场景中,其累计的平均直方图的形状是与特定场景相关的。提取高斯拟合后场均直方图各峰的峰高、峰位和峰宽等特征,利用最小二乘法,提出了一种基于多颜色空间和场均直方图的视频场景分类方法,并以常见的五种视频场景为例进行了初步应用。实验表明,该方法具有较好的分类效果。     

一种面向空间数据的聚集查询处理方法

提出一种新的面向空间数据的聚集查询处理方法。该方法扩展了AMH和AMH*方法,首先将整个空间栅格化,再将若干频数接近的邻居栅格组合成遵循一定约束条件的桶。随着数据分布变化,各桶不断重组或者拆分,具有空间复杂度低、查询精度高等特点。理论分析与实验结果均表明新方法优于现有方法。