Morphogenesis of Floating Bone Segments: A Legacy of Serial Tensile Cross-Strut Microdamage in Trabecular Disconnection “Crumple Zones”?

Trabecular bone disconnection “hotspots” of real termini (ReTm) previously mapped as loci of weakness in the female aging spine and hip may be a source of free-floating cancellous segments found in the medullary space using a bespoke, thick slice histological method for identifying ReTm. A factor in their origin is apparently microdamage proliferation (differentiated by en bloc silver staining) with occasional callus moderation. Validation of similar “floating segments” (FS) in the ex-breeder rat suggested a pilot model for a potentially common phenomenon. Following marrow elution and density fractionation of the isolated floating segments from the whole proximal rat femora, scanning electron microscopy (SEM) and elemental microanalysis (EDS) was performed. The eluent contained numbers of vertically truncated, laterally branched floating segments (acute severance of sequential tensile cross-struts, causing chronic compression overload of axial-struts, with ii) inadequate stabilising callus, facilitating ReTm stacking into predetermined, substructural “crumple zones” of force containment, spheroidal attrition and particulate dissociation. As a catabolic outcome of altered tensile and hormonal influence, FS number may add a novel variable to cancellous bone kinetics particularly in women of relevance to fracture predisposition.

Influence of lens position as detected by an anterior segment analysis system on postoperative refraction in cataract surgery

AIM:To predict postoperative intraocular lens(IOL)position using the Sirius anterior segment analysis system and investigate the effect of lens position and IOL type on postoperative refraction.METHODS:A total of 97 patients(102 eyes)were enrolled in the final analysis.An anterior segment biometry measurement was performed preoperatively with Sirius and Lenstar.The results of predicted lens position(PLP)and IOL power were automatically calculated by the software used by the instruments.Effective lens position(ELP)was measured manually using Sirius 3 mo postoperatively.Pearson's correlation analysis and linear regression analysis were used to determine the correlation of lens position to other parameters.RESULTS:PLP and ELP were positively correlated to axial length(AL;r=0.42,P<0.0001 and r=0.49,P<0.0001,respectively).There was a weak correlation between the peLP(ELP-PLP)and the prediction error of spherical refraction(peSR;r=0.34,P<0.0001).The peLP of Softec HD IOL differed statistically from those of both the TECNIS ZCB00 and Sensor AR40E IOLs.Multiple linear regression was used to obtain the prediction formula:ELP=0.66+0.63×[aqueous depth(AQD)+0.6 LT](r=0.61,P<0.0001),and a new variable(AQD+0.6 LT)was found to have the strongest correlation with ELP.CONCLUSION:The Sirius anterior segment analysis system is helpful to predict ELP,which reduces postoperative refraction error.

Slip rate and interseismic coupling along the Moxi-Shimian segment of the Xianshuihe fault, Eastern Qinghai-Xizang Plateau

On September 5,2022,the Luding M6.8 earthquake occurred in the Moxi-Shimian segment of the Xianshuihe fault,coinciding with the historical ruptured zone of the 1786 Moxi earthquake.Its seis-mogenic environment provides a foundation for comprehending the mechanism of the earthquake and its future hazard.In the Moxi-Shimian segment,we establish a series of near-field Global Navigation Satellite Systems(GNsS)stations to enhance the spatial resolution of observational data for the inversion of the interseismic kinematic parameters.In this study,with an elastic screw dislocation model con-strained by GNsS observations,the slip rate of the Moxi-Shimian segment is estimated to be 10.9±1.0 mm/yr,while the locking depth is 15.7±6.2 km.Additionally,we utilize a block-dislocation model to invert the interseismic fault coupling along the Kangding-Moxi-Shimian segment.The result indicates a gradual deepening of the locking depth along the section from Kangding to Shimian.The coseismic rupture of the 2022 event occurred within the high coupling regions in the Kangding-Moxi-Shimian segment,which indicates that the rupture kinematics in this event might be controlled by the interseismic deformation.The seismic moment accumulated within the ruptured zone of the Luding earthquake since 1786 ranges in[1.42-3.40]×10^(19) N-m,which is significantly greater than the seismic moment released during the 2022 event.As a result,we infer that the Luding earthquake released only a portion of the accumulated energy within the original rupture zone since 1786,indicating that the 2022 event has not caused a complete rupture in the Moxi-Shimian segment.Consequently,there remains a substantial seismic hazard in this area.

A New Approach to Simulate Positional Error of Line Segment in GIS

To determine the distribution of positional error of a line segment, Monte Carlo approach is applied to simulate the probability density function of a line segment with the assumption that the error of endpoints in a line segment follows a two-dimensional normal distribution. For such purpose, a stochastic generator used for uncertain endpoints with the two-dimensional normal distribution is presented. This forms the basis of the generation of random line segment for the simulation of the error model of a whole line segment. The error models cover the cases where two endpoints are either independent or dependent to each other, also including a special case that the distance between two random endpoints in a line segment is close enough.

Left jackknife position:a novel position for laparoscopic hepatectomy

Background: Laparoscopic hepatectomy for hepatocellular carcinoma(HCC) located in segment Ⅵ, Ⅶ, or Ⅷ of the liver is usually difficult because of poor operative exposure, due to the unique anatomical structure. In this study, we evaluated the practice of laparoscopic hepatectomy with the left jackknife position for patients with HCC located in segment Ⅵ, Ⅶ, or Ⅷ.Methods: A total of 10 patients were enrolled to undergo laparoscopic hepatectomy with the left jackknife position.Tumors located in segment Ⅵ, Ⅶ, or Ⅷ were assessed by preoperative dynamic computed tomography or magnetic resonance imaging. Operation time, intraoperative blood loss, postoperative fasting time, postoperative drainage time, major postoperative complications, and duration of postoperative hospital stay were recorded.Results: All surgeries were successfully completed. None of the patients required conversion to open surgery during the procedure, and no serious postoperative complications were observed.The median tumor size was 31 mm(range 23-41 mm) in diameter, the mean operation time was 166 ± 38 min, the mean intraoperative blood loss was220 ± 135 mL, and the median postoperative hospital stay was 4 days(range 2-7 days).Conclusions: For HCC located in segment Ⅵ, Ⅶ, or Ⅷ, laparoscopic hepatectomy with this novel position—the left jackknife position—is safe and effective during tumor resection by exposing a sufficient operating field.Trial registration ClinicalTrials.gov ID:

A hybrid attention deep learning network for refined segmentation of cracks from shield tunnel lining images

This research developed a hybrid position-channel network (named PCNet) through incorporating newly designed channel and position attention modules into U-Net to alleviate the crack discontinuity problem in channel and spatial dimensions. In PCNet, the U-Net is used as a baseline to extract informative spatial and channel-wise features from shield tunnel lining crack images. A channel and a position attention module are designed and embedded after each convolution layer of U-Net to model the feature interdependencies in channel and spatial dimensions. These attention modules can make the U-Net adaptively integrate local crack features with their global dependencies. Experiments were conducted utilizing the dataset based on the images from Shanghai metro shield tunnels. The results validate the effectiveness of the designed channel and position attention modules, since they can individually increase balanced accuracy (BA) by 11.25% and 12.95%, intersection over union (IoU) by 10.79% and 11.83%, and F1 score by 9.96% and 10.63%, respectively. In comparison with the state-of-the-art models (i.e. LinkNet, PSPNet, U-Net, PANet, and Mask R–CNN) on the testing dataset, the proposed PCNet outperforms others with an improvement of BA, IoU, and F1 score owing to the implementation of the channel and position attention modules. These evaluation metrics indicate that the proposed PCNet presents refined crack segmentation with improved performance and is a practicable approach to segment shield tunnel lining cracks in field practice.

Effect of continuous positive airway pressure ventilation on nocturnal ST-segment changes in patients with sleep-disordered breathing

Objective To determine whether sleep-disordered breathing (SDB) may lead to nocturnal myocardial ischemia and whether the severity of this ischemia may be relieved by nasal continuous positive airway pressure (CPAP). Methods Overnight polysomnogram examination and simultaneous 3-channel Holter monitoring were performed on 76 patients with moderate to severe SDB and no history of coronary heart disease. All the cases were treated with CPAP for one night. ST depression was defined as a ST segment decrease of more than 1 mm from baseline and lasting 1 min or more. The total duration (minutes) of ST depression was indexed to the total sleep time (minutes per hour of sleep). Results Twenty-eight patients (37%) showed ST segment depression during their sleep. Before CPAP treatment, the respiratory disturbance index (RDI) and arousal index were significantly higher during periods of ST depression than when ST segments were isoelectric, whereas no significant difference was found in blood oxygen saturation (SaO2). After the CPAP treatment of patients with ST depression, the duration of ST depression was significantly reduced from 36.8±18.9 to 11.4±13.2 min/h (P<0.05). ST depression-related indexes, including RDI, arousal index and the percentage of sleep time spent at SaO2 below 90% (TS90/ TST), were all significantly decreased, with RDI from 63.4±23.8 to 8.1±6.6 /h, arousal index from 51.2±18.9 to 9.6±5.4 /h, and TS90/ TST from 50.6±21.4 to 12.9±14.7% (P<0.05). Conclusion ST-segment depression is rather common in patients with moderate to severe SDB, and CPAP treatment can significantly reduce the duration of ST depression. ST depression in these patients may reflect the myocardial ischemia that really exists and the non-ischemic changes associated with recurrent SDB.

Sea-Land Segmentation of Remote Sensing Images Based on SDW-UNet

Image segmentation of sea-land remote sensing images is of great importance for downstream applications including shoreline extraction,the monitoring of near-shore marine environment,and near-shore target recognition.To mitigate large number of parameters and improve the segmentation accuracy,we propose a new Squeeze-Depth-Wise UNet(SDW-UNet)deep learning model for sea-land remote sensing image segmentation.The proposed SDW-UNet model leverages the squeeze-excitation and depth-wise separable convolution to construct new convolution modules,which enhance the model capacity in combining multiple channels and reduces the model parameters.We further explore the effect of position-encoded information in NLP(Natural Language Processing)domain on sea-land segmentation task.We have conducted extensive experiments to compare the proposed network with the mainstream segmentation network in terms of accuracy,the number of parameters and the time cost for prediction.The test results on remote sensing data sets of Guam,Okinawa,Taiwan China,San Diego,and Diego Garcia demonstrate the effectiveness of SDW-UNet in recognizing different types of sea-land areas with a smaller number of parameters,reduces prediction time cost and improves performance over other mainstream segmentation models.We also show that the position encoding can further improve the accuracy of model segmentation.

阅读伴随词汇学习的词切分:首、尾词素位置概率的不同作用

本研究通过两个平行实验,探讨重复学习新词时首、尾词素位置概率信息作用于词切分的变化模式。采用阅读伴随词汇学习范式,将双字假词作为新词,实验1操纵首词素位置概率高低,保证尾词素相同;实验2操纵尾词素位置概率高低,保证首词素相同。采用眼动仪记录大学生阅读时的眼动轨迹。结果显示:(1)首、尾词素位置概率信息的词切分作用随新词在阅读中学习次数的增加而逐步变小,表现出“熟悉性效应”。(2)首词素位置概率信息的“熟悉性效应”表现在回视路径时间、总注视次数两个相对晚期的眼动指标,而尾词素位置概率信息的“熟悉性效应”则从凝视时间开始,到回视路径时间,再持续到总注视时间。结果表明首、尾词素的位置概率信息均作用于阅读伴随词汇学习的词切分,但首词素的作用时程更长,更稳定,支持了首词素在双字词加工中具有优势的观点。

基于形色筛选的苹果园羽化害虫粘连图像分割方法

针对苹果园害虫识别过程中的粘连问题,提出了一种基于形色筛选的害虫粘连图像分割方法。首先,采集苹果园害虫图像,聚焦于羽化害虫。害虫在羽化过程中已完成大部分生长发育,其外部形态、颜色、纹理更为稳定显著。因此,基于不同种类害虫的形色特征信息分析,来获取害虫HSV分割阈值和模板轮廓。其次,利用形状因子判定分割粘连区域,通过颜色分割法和轮廓定位分割法来实现非种间与种间粘连害虫的分割。最后,对采集的苹果园害虫图像进行了试验分析,采用基于形色筛选的分割法对单个害虫进行分割,结果表明,本文方法的平均分割率、平均分割错误率和平均分割有效率分别为101%、3.14%和96.86%,分割效果优于传统图像分割方法。此外,通过预定义的颜色阈值,本文方法实现了棉铃虫、桃蛀螟与玉米螟的精准分类,平均分类准确率分别为97.77%、96.75%与96.83%。同时,以Mask R-CNN模型作为识别模型,平均识别精度作为评价指标,分别对已用本文方法和未用本文方法分割的害虫图像进行识别试验。结果表明,已用本文方法分割的棉铃虫、桃蛀螟和玉米螟害虫图像平均识别精度分别为96.55%、94.80%与95.51%,平均识别精度分别提高16.42、16.59、16.46个百分点。这表明该方法可为果园害虫精准识别提供理论和方法基础。

连栋温室分段变距喷雾机器人设计与试验

针对国内连栋温室缺乏植保喷雾机、机械走直定位与换轨转向精度低等问题,设计了一种连栋温室分段变距喷雾机器人,在实现无人化喷药的同时提高作业精度。为满足连栋温室机械作业路轨结合、精准切换的要求,提出一种通用型移动底盘,并确定其关键设计参数;为减少底盘上下轨时的偏移量,设计轨上矫正装置,通过分析计算及试验验证,确定其安装余量为4 mm;针对底盘对轨误差大的问题,提出一种二维码融合陀螺仪及光电传感器双向垂直寻迹的路面关键点定位与转向控制方法。设计分段变距喷雾装置,提出一种丝杆滑台驱动的喷杆变距方案,分析校验其驱动参数以满足工作要求;基于滚针轴承设计喷杆辅助防抖装置,减小因喷杆剧烈抖动带来的滑台与喷杆损伤。开发底盘运动及分段变距喷雾控制系统,实现喷雾机器人在连栋温室内的全程自动化作业。最后,对样机进行底盘性能与喷雾效果试验。底盘作业时直线行走与对轨误差平均值分别为4.8、5.8 mm,满足控制精度要求;避障距离为34 cm,满足安全性要求;防抖装置的安装使喷杆行进方向的抖动量从-1°~1.3°降低到±0.4°内,喷头方向的抖动量从±0.5°降低到±0.3°内,防抖效果显著;分段变距喷雾作业后,盛果期番茄叶片正面雾滴沉积量为1.76μL/cm^(2),反面沉积量为0.2μL/cm^(2),雾滴体积中径在100~180μm之间,满足作业要求。

基于点云特征的改进RANSAC地面分割算法

针对室外复杂场景下,轻量级和地面优化的激光雷达里程计与测图(LeGO-LOAM)算法由于地面分割不精确而导致算法定位精度降低的问题,提出一种基于改进随机一致性采样(RANSAC)的多线程地面分割算法:相较于传统RANSAC算法,该算法舍弃从全部原始数据中随机选取种子点拟合地面模型的迭代方式,首先利用点云高程、曲率等点特征信息挑选出所有小于高程、曲率等阈值的种子点以构建种子点集合,并根据种子点集合中的种子点数量判断是否需要多线程处理;然后根据判断结果从种子点集合中选择种子点子集进行地面拟合;最后比较各地面模型所包含的点云数量以获得最优地面模型参数以及地面点云集;地面分割精度的提高有效地降低了LeGO-LOAM算法的定位误差。实验结果表明,在室外复杂场景下所提出的地面分割算法分割效果更好,杂点更少;相较于原LeGO-LOAM算法,改进算法的定位误差降低至3.73 m,平面均方根误差降低了20.8%。

基于传感控制的自动采棉机系统设计

为了实现棉花依据成熟度进行采摘,基于图像传感器设计图像处理方法,完成棉桃图像特征提取,进而实现棉花成熟度评估和空间棉桃定位。采用HSV方案进行图像灰度处理,采用二维大津法进行图像分割,提取棉桃特征参数。对成熟棉花和未成熟棉花图像样本分析发现,棉桃图像轮廓周长L和面积S比大于0.31时棉花成熟,从而建立棉花采摘标准,并采用双目图像系统,实现目标棉桃的图像坐标向空间坐标的转换。对图像分割精度,棉桃提取精度和视觉定位精度进行测试,结果表明:系统具有较高的可靠性。

融合位置注意力机制与轻量化STDC网络的非结构化场景语义分割

近年来,非结构化道路分割已成为计算机视觉领域的重要研究方向之一.现有的大多数方法适合结构化道路的分割并无法满足非结构化道路分割的准确性与实时性需求.为了解决上述问题,本文对STDC网络进行改进,引入残差连接来更好地融合多尺度语义信息,还提出一种嵌入位置注意力模块的空洞空间卷积池化金字塔(PAASPP)来增强网络对道路等特定区域的位置感知能力.本文在RUGD与RELLIS-3D两个数据集上进行实验,所提出方法的MIoU在两个数据集的测试集上分别达到了50.78%和49.96%.

融合双重极化注意力的轻量化半监督语义分割

针对目前半监督语义分割方法复杂度高、训练精度低、参数量过大等问题,提出融合双重极化自注意力机制的轻量级半监督语义分割算法。模型使用由位置感知循环卷积构造的Resnet-101残差网络作为分割骨干网络提取深层特征。融合了通道及空间双重极化自注意力机制,在极化通道和空间注意力分支中保持较高内部分辨率。将位置感知循环卷积与通道注意力操作结合起来,提升分割精度并降低计算成本,克服硬件支持等问题。在公开数据集PASCALVOC 2012上的实验结果显示,该算法其平均交并比可达到76.32%,较基准模型准确率提高了2.52个百分点,参数量减少了9%,模型硬件所占内存减小了61.6%。设计的模型与领域内最新算法相比,该算法在精度、模型复杂度、参数量等方面均展现出了显著的优势。

基于定位观测站图像实现草原植被覆盖度自动化提取的方法比较

植被覆盖度是草原生态监测和研究中极其重要的指标。本研究采用无人值守定位观测站获取样方图像数据集,比较选取多种提取算法,旨在解决定位连续观测中不同样方的大量图像数据集如何自动化提取植被覆盖度指标的难题。本文使用绿度指数法、随机森林(RF)、支持向量机(SVM)和反向传播(BP)神经网络不同图像分割方法提取不同草地类型的覆盖度,并讨论绿度指数之类的阈值分割方法与机器学习方法的优劣、3种机器学习算法产生不同分类效果的原因以及覆盖度数值产生误差的主要原因。结果表明,机器学习算法可灵活解决定位观测站样方图像集的快速自动化提取植被覆盖度问题。绿度指数阈值分割方法应用于植被覆盖度的分割效果较差;RF算法在高寒草原的分割精度较高,SVM在温性草原和温性荒漠草原中分割效果较高,BP神经网络在高寒草甸的覆盖度提取中更有优势。本研究可为新时代草原生态监测的信息化和智能化监测设备研发提供重要的参考依据。

TBM盾构管片台柱型定位销设计制备技术研究及工程应用

全断面硬岩隧道掘进(TBM)施工过程中同一断面内的相邻两个衬砌管片之间需要进行连接,若拼接施工存在问题或定位销出现变形,会引起盾构管片的上浮和错台。文章分析了传统定位销存在的缺点及施工方面的不便,设计了一种新型的定位销,对管片模具及预埋方式进行了修改,并对定位销的承载性能进行了研究。研究结果表明,将原来的纺锤形结构改为台柱结构,设置有多圈凸纹,可以有效增大定位销与定位销孔之间的摩擦力,方便施工和便捷安装。控制注塑材料的抗拉强度不低于80 MPa,制备而成的台柱定位销剪切强度达到了420.0 MPa以上,具有很高的力学性能。所制备的定位销在某引水隧洞工程中进行了工程应用,降低了TBM管片的制作难度及定位销的安装难度,取得了良好的工程效果。

基于位置信息和注意力机制的路面裂缝检测

路面裂缝是造成公路安全问题的主要因素。传统的裂缝检测通常以人工检测为主,存在效率低、不安全等问题,此外现有深度学习检测模型在面临阴影遮挡、背景复杂等干扰因素时会造成裂缝检测不完整。针对上述问题,提出一种基于位置信息和注意力机制的路面裂缝检测模型(PA-TransUNet)。首先,通过混合编码器接收输入图像,提取裂缝特征信息,引入查询项、键、值的位置信息,提升编码器Transformer中自注意力机制捕获裂缝形状和补偿特征信息丢失的能力。然后,输入裂缝特征到解码器进行上采样,设计一种基于注意力门控的解码模块(AGDM),AGDM通过抑制非裂缝区域来加强对裂缝区域的学习,提高裂缝检测的准确性和完整性。实验结果表明,PA-TransUNet模型在路面裂缝检测数据集(CFD)和Cracktree200这2个公开数据集上的F1值分别达到87.44%和82.58%。此外,为了进一步检验PA-TransUNet模型在实际工程中的裂缝检测能力,又在自制无人机裂缝(UAV Cracks)数据集上取得了88.68%的F1值,由此可见其能较好地满足实际工程中的裂缝检测需求。

“E骨”-反肩置换术一站式术前规划系统

目的为提高反肩置换术术前规划的精准度和效率,本研究开发了“E骨”—反肩置换术一站式术前规划系统。方法利用nnU-net深度神经网络进行肩胛骨分割,获得精确的肩胛骨分割结果;结合骨密度、上下倾角和入钉长度3个关键因素自动定位基座,并计算出手术规划所需定量参数;结合关节盂形态和基座定位信息生成个性化导板;编写系统界面,将各部分功能模块化组装便于使用,提供交互操作与规划结果显示的功能。结果构建出“E骨”术前规划系统。本系统与Mimics系统相比,减少了规划过程中繁琐的手动调整,平均入钉长度规划结果长于Mimics系统,且规划时间缩短86%。本系统肩胛骨分割精度达到99.93%,较Mimics更高。结论本研究为反肩置换术提供了一套一站式的高效、精准的术前规划方案,具有广泛的临床应用前景。

基于高光谱成像的烤烟着生部位识别

【目的】采用高光谱成像技术结合机器学习方法,建立烤烟着生部位(上部、中部、下部)的识别模型。【方法】首先,通过分析烟叶在水、氮敏感波段下的强度分布特征,采用了一种结合OTSU和Sauvola图像分割算法的双阈值感兴趣区(ROI)选取方法,然后对比分析不同预处理方法对数据建模的影响规律,采用支持向量机(SVM)、极限梯度提升(XGBoost)算法进行判别模型的建立,通过参数寻优进行模型的优化。使用遗传算法(GA)和遗传算法结合连续投影算法(GA-SPA)进行特征波长的选择,建立简化模型。【结果】(1)建立的双阈值感兴趣区选取方法能准确高效地实现烤烟叶片正常叶面区域的选取(2)不同数据预处理方法对识别模型影响较为显著,基于一阶导和萨维莱茨-戈莱平滑(1Der+SG)预处理光谱数据,结合GA选取的特征波长建立的XGBoost着生部位识别模型具有最佳的分类效能,其准确率高达97.78%。【结论】研究建立的基于高光谱成像技术结合机器学习方法的部位模型可满足烤烟着生部位的高效准确识别。