Anisotropy of Trabecular Bone from Ultra-Distal Radius Digital X-Ray Imaging: Effects on Bone Mineral Density and Age

Background: When applied to trabecular bone X-ray images, the anisotropic properties of trabeculae located at ultra-distal radius were investigated by using the trabecular bone scores (TBS) calculated along directions parallel and perpendicular to the forearm. Methodology: Data from more than two hundred subjects were studied retrospectively. A DXA (GE Lunar Prodigy) scan of the forearm was performed on each subject to measure the bone mineral density (BMD) value at the location of ultra-distal radius, and an X-ray digital image of the same forearm was taken on the same day. The values of trabecular bone score along the direction perpendicular to the forearm, TBSx, and along the direction parallel to the forearm, TBSy, were calculated respectively. The statistics of TBSx and TBSy were calculated, and the anisotropy of the trabecular bone, which was defined as the ratio of TBSy to TBSx and changed with subjects’ BMD and age, was reported and analyzed. Results: The results show that the correlation coefficient between TBSx and TBSy was 0.72 (p BMD and age was reported. The results showed that decreased trabecular bone anisotropy was associated with deceased BMD and increased age in the subject group. Conclusions: This study shows that decreased trabecular bone anisotropy was associated with decreased BMD and increased age.

COVID-19 Detection from Chest X-Ray Images Using Convolutional Neural Network Approach

COVID-19 is a respiratory illness caused by the SARS-CoV-2 virus, first identified in 2019. The primary mode of transmission is through respiratory droplets when an infected person coughs or sneezes. Symptoms can range from mild to severe, and timely diagnosis is crucial for effective treatment. Chest X-Ray imaging is one diagnostic tool used for COVID-19, and a Convolutional Neural Network (CNN) is a popular technique for image classification. In this study, we proposed a CNN-based approach for detecting COVID-19 in chest X-Ray images. The model was trained on a dataset containing both COVID-19 positive and negative cases and evaluated on a separate test dataset to measure its accuracy. Our results indicated that the CNN approach could accurately detect COVID-19 in chest X-Ray images, with an overall accuracy of 97%. This approach could potentially serve as an early diagnostic tool to reduce the spread of the virus.

Electromagnetic scattering and imaging simulation of extremely large-scale sea-ship scene based on GPU parallel technology

Aiming to solve the bottleneck problem of electromagnetic scattering simulation in the scenes of extremely large-scale seas and ships,a high-frequency method by using graphics processing unit(GPU)parallel acceleration technique is proposed.For the implementation of different electromagnetic methods of physical optics(PO),shooting and bouncing ray(SBR),and physical theory of diffraction(PTD),a parallel computing scheme based on the CPU-GPU parallel computing scheme is realized to balance computing tasks.Finally,a multi-GPU framework is further proposed to solve the computational difficulty caused by the massive number of ray tubes in the ray tracing process.By using the established simulation platform,signals of ships at different seas are simulated and their images are achieved as well.It is shown that the higher sea states degrade the averaged peak signal-to-noise ratio(PSNR)of radar image.

How to Coadd Images.Ⅱ.Anti-aliasing and PSF Deconvolution

We have developed a novel method for co-adding multiple under-sampled images that combines the iteratively reweighted least squares and divide-and-conquer algorithms.Our approach not only allows for the anti-aliasing of the images but also enables Point-Spread Function(PSF)deconvolution,resulting in enhanced restoration of extended sources,the highest peak signal-to-noise ratio,and reduced ringing artefacts.To test our method,we conducted numerical simulations that replicated observation runs of the China Space Station Telescope/the VLT Survey Telescope(VST)and compared our results to those obtained using previous algorithms.The simulation showed that our method outperforms previous approaches in several ways,such as restoring the profile of extended sources and minimizing ringing artefacts.Additionally,because our method relies on the inherent advantages of least squares fitting,it is more versatile and does not depend on the local uniformity hypothesis for the PSF.However,the new method consumes much more computation than the other approaches.

M^(3)Res-Transformer:新冠肺炎胸部X-ray图像识别模型

新冠肺炎(COVID-19)自爆发以来严重影响人类生命健康,近年来残差神经网络广泛应用于COVID-19识别任务中,辅助医生快速地诊断COVID-19患者,但是COVID-19图像病变区域形状复杂、大小不一,与周围组织的边界模糊,导致网络难以提取有效特征.本文针对上述问题,提出一种M^(3)Res-Transformer的新冠肺炎胸部X-ray图像识别模型,采用Res-Transformer作为模型的主干网络,结合ResNet和ViT,有效地整合局部病变特征和全局特征;设计混合残差注意力模块(mixed residual attention Module,mraM),同时考虑通道和空间位置的相互依赖性,增强网络的特征表达能力;为了增大感受野,提取多尺度特征,通过叠加具有不同扩张率的扩张卷积构造多尺度扩张残差模块(multiscale dilated residual Module,mdrM),根据不同层次特征尺度的差异,使用3个逐渐收缩尺度的mdrM进行多尺度特征提取;提出上下文交叉感知模块(contextual cross-awareness Module,ccaM),使用深层特征中的语义信息来引导浅层特征,然后将浅层特征中的空间信息嵌入深层特征中,采用交叉加权注意力机制高效聚合深层和浅层特征,获得更丰富的上下文信息.为了验证本文所提模型的有效性,在新冠肺炎胸部X-ray图像数据集上进行实验,与先进的CNN分类模型、融合不同注意力机制的ResNet50模型、基于Transformer的分类模型对比以及消融实验.结果表明,本文所提模型的Acc、Pre、Rec、F1-Score与Spe指标分别为96.33%、96.36%、96.33%、96.35%与96.26%,在COVID-19胸部X-ray图像识别任务中有效提升了识别精度,并通过可视化方法对其进行进一步验证,为COVID-19的辅助诊断提供重要的参考价值.

ZnSb/Ti_(3)C_(2)T_(x)MXene van der Waals heterojunction for flexible near-infrared photodetector arrays

Two-dimension(2D)van der Waals heterojunction holds essential promise in achieving high-performance flexible near-infrared(NIR)photodetector.Here,we report the successful fabrication of ZnSb/Ti_(3)C_(2)T_(x)MXene based flexible NIR photodetector array via a facile photolithography technology.The single ZnSb/Ti_(3)C_(2)T_(x)photodetector exhibited a high light-to-dark current ratio of 4.98,fast response/recovery time(2.5/1.3 s)and excellent stability due to the tight connection between 2D ZnSb nanoplates and 2D Ti_(3)C_(2)T_(x)MXene nanoflakes,and the formed 2D van der Waals heterojunction.Thin polyethylene terephthalate(PET)substrate enables the ZnSb/Ti_(3)C_(2)T_(x)photodetector withstand bending such that stable photoelectrical properties with non-obvious change were maintained over 5000 bending cycles.Moreover,the ZnSb/Ti_(3)C_(2)T_(x)photodetectors were integrated into a 26×5 device array,realizing a NIR image sensing application.

面向牙齿X-RAY图像分割的牙齿模型

医学图像分割技术为临床诊断和治疗提供关键依据,但传统的图像分割技术易将图像过度分割且分割效果不够精细。为解决这些问题,本文研究采用U-Net网络分割牙齿X-Ray图像,通过合理选择池化操作、激活函数和周期数目,解决牙齿X-Ray图像中存在的牙齿与周围组织的对比度低、边界模糊、牙齿与背景分布不均、牙齿与组织粘连等问题。实验结果表明,U-Net网络具备有效的牙齿X-Ray图像的分割性能。

Comparison of efficacy of lung ultrasound and chest X-ray in diagnosing pulmonary edema and pleural effusion in ICU patients: A single centre, prospective, observational study

Background and Aims While chest X-ray (CXR) has been a conventional tool in intensive care units (ICUs) to identify lung pathologies, computed tomography (CT) scan remains the gold standard. Use of lung ultrasound (LUS) in resource-rich ICUs is still under investigation. The present study compares the utility of LUS to that of CXR in identifying pulmonary edema and pleural effusion in ICU patients. In addition, consolidation and pneumothorax were analyzed as secondary outcome measures. Material and Methods This is a prospective, single centric, observational study. Patients admitted in ICU were examined for lung pathologies, using LUS by a trained intensivist;and CXR done within 4 hours of each other. The final diagnosis was ascertained by an independent senior radiologist, based on the complete medical chart including clinical findings and the results of thoracic CT, if available. The results were compared and analyzed. Results Sensitivity, specificity and diagnostic accuracy of LUS was 95%, 94.4%, 94.67% for pleural effusion;and 98.33%, 97.78%, 98.00% for pulmonary edema respectively. Corresponding values with CXR were 48.33%, 76.67%, 65.33% for pleural effusion;and 36.67%, 82.22% and 64.00% for pulmonary edema respectively. Sensitivity, specificity and diagnostic accuracy of LUS was 91.30%, 96.85%, 96.00% for consolidation;and 100.00%, 79.02%, 80.00% for pneumothorax respectively. Corresponding values with CXR were 60.87%, 81.10%, 78.00% for consolidation;and 71.3%, 97.20%, 96.00% for pneumothorax respectively. Conclusion LUS has better diagnostic accuracy in diagnosis of pleural effusion and pulmonary edema when compared with CXR and is thus recommended as an effective alternative for diagnosis of these conditions in acute care settings. Our study recommends that a thoracic CT scan can be avoided in most of such cases.

Application of Dual-Energy X-Ray Image Detection of Dangerous Goods Based on YOLOv7

X-ray security equipment is currently a more commonly used dangerous goods detection tool, due to the increasing security work tasks, the use of target detection technology to assist security personnel to carry out work has become an inevitable trend. With the development of deep learning, object detection technology is becoming more and more mature, and object detection framework based on convolutional neural networks has been widely used in industrial, medical and military fields. In order to improve the efficiency of security staff, reduce the risk of dangerous goods missed detection. Based on the data collected in X-ray security equipment, this paper uses a method of inserting dangerous goods into an empty package to balance all kinds of dangerous goods data and expand the data set. The high-low energy images are combined using the high-low energy feature fusion method. Finally, the dangerous goods target detection technology based on the YOLOv7 model is used for model training. After the introduction of the above method, the detection accuracy is improved by 6% compared with the direct use of the original data set for detection, and the speed is 93FPS, which can meet the requirements of the online security system, greatly improve the work efficiency of security personnel, and eliminate the security risks caused by missed detection.

基于X-ray和RGB图像融合的实蝇侵染柑橘无损检测

实蝇侵染柑橘流入市场会造成巨大的经济损失,因此需要在商品化处理阶段对其全面筛除。针对柑橘在实蝇侵染早期没有明显外部特征,人工抽样检测效率低、筛除难的问题,探索了在生产线上同时搭载农业X光机与RGB相机进行无损检测的可行性,提出了基于X-ray(X光)和RGB图像的多模态数据融合方法,建立了CNN-LSTM检测模型,实现了实蝇侵染柑橘高精度无损检测。模拟了柑橘在生产线上滚动并被拍摄6幅X-ray和RGB序列图像的过程,构建了实蝇侵染柑橘的多源数据集,融合了不同模态的实蝇侵染特征信息,提升了实蝇侵染柑橘检测模型的检测能力,并对比了ResNet18-LSTM、GoogleNet-LSTM、SqueezeNet-LSTM、MobileNetV2-LSTM轻量化检测模型,验证了多模态数据融合方法的有效性。研究结果表明,提出的多模态数据融合实蝇侵染柑橘方法比单模态检测方法检测性能更加优异,其中ResNet18-LSTM检测准确率最高,多模态的图像融合和特征融合方法检测准确率分别达到97.3%和95.7%,单模态X-ray和RGB检测方法准确率分别为93.2%和89.3%。本研究可为实蝇侵染柑橘在线无损检测技术与装备的研究提供理论支撑。

A Novel Method for Automated Lung Region Segmentation in Chest X-Ray Images

Detecting and segmenting the lung regions in chest X-ray images is an important part in artificial intelligence-based computer-aided diagnosis/detection (AI-CAD) systems for chest radiography. However, if the chest X-ray images themselves are used as training data for the AI-CAD system, the system might learn the irrelevant image-based information resulting in the decrease of system’s performance. In this study, we propose a lung region segmentation method that can automatically remove the shoulder and scapula regions, mediastinum, and diaphragm regions in advance from various chest X-ray images to be used as learning data. The proposed method consists of three main steps. First, employ the simple linear iterative clustering algorithm, the lazy snapping technique and local entropy filter to generate an entropy map. Second, apply morphological operations to the entropy map to obtain a lung mask. Third, perform automated segmentation of the lung field using the obtained mask. A total of 30 images were used for the experiments. In order to verify the effectiveness of the proposed method, two other texture maps, namely, the maps created from the standard deviation filtering and the range filtering, were used for comparison. As a result, the proposed method using the entropy map was able to appropriately remove the unnecessary regions. In addition, this method was able to remove the markers present in the image, but the other two methods could not. The experimental results have revealed that our proposed method is a highly generalizable and useful algorithm. We believe that this method might act an important role to enhance the performance of AI-CAD systems for chest X-ray images.

基于RetinaNet-AACIDD的铝合金铸件X-ray图像缺陷检测方法

为了解决铝合金铸件X射线图像噪声多、缺陷背景复杂,缺陷较难检测的问题,制作了一个含有14640张准确标注的大型铝合金铸件内部X射线图像缺陷数据集ALU-Xray,提出一种基于深度学习的缺陷检测算法RetinaNet-AACIDD(RetinaNet for aluminum alloy casting internal defect detection)。通过加入由通道注意力模块C-Block和空间注意力模块S-Block组成的混合注意力模块CS-Block,有效降低了噪声、背景等无关信息对检测精度的干扰,并通过多尺度特征融合模块,完成了深层语义信息的向下传递,结合多尺度特征预测和基于改进的聚类算法重新设计的锚框,使得模型对气泡、裂纹等尺度较小的缺陷和复杂背景下的疏松、缩孔等缺陷的检测精度显著提升。

一种新的术中X线与术前CT图像配准方法

目的本研究旨在配准胸主动脉血管内修复术(thoracic endovascular aortic repair,TEVAR)术中X线与术前CT图像,为TEVAR支架植入提供精确安全的导航。然而,现有配准算法存在无法有效弥合投影CT图像生成的数字重建影像(digitally reconstructed radiography,DRR)与X线图像之间的域间差异和难以获得图像分割标签的问题。因此,需要提出新的方法来改善这一问题。方法本文提出了一种新的配准框架,该框架结合了基于生成对抗网络(generative adversarial network,GAN)的域自适应网络和基于Transformer的配准网络。基于GAN的域自适应网络将X线图像的风格迁移到DRR图像上,使两者在图像风格上更接近。基于Transformer的配准网络采用CNN与跨模态变换器(cross-modality transformer,CMT)相结合的模式,直接配准X线与CT图像,无需进行图像分割。结果本文在208对标定的TEVAR术中X线与CT图像对上对新的配准方法进行了验证。与其他域适应方法相比,本文所采用的CycleGAN网络作为风格转换模块,有效减小了DRR图像与X线图像之间的域间差异。消融实验结果进一步证实,配准网络中的全局局部感知模块(global-local perception module,GLPM)对提高配准精度具有明显作用,而空间缩减(spatial reduction,SR)则有效缩短了配准时间。通过对比现有方法和本文方法在真实患者X线与CT图像对上的配准效果,本文的方法在配准精度和成功率方面均表现出最佳性能。结论本文提出的新的X线与CT图像配准方法有效克服了现有方法存在的域间差异以及难以获得分割标签的问题。

基于迁移学习与残差网络的快递包裹X光图像识别

针对快递包裹违禁物品识别存在种类繁多、依赖人力和X光图像获取难度大等问题,为提高快递包裹违禁物品识别的效率和准确度,本研究提出一种迁移学习与残差网络相结合的快递包裹X光图像识别方法(TL-ResNet18)。首先构建了相似度高的源领域数据集和目标领域数据集;其次,选用ResNet18作为预训练模型,调整初始化参数结构,并将ResNet18学习到的内容作为初始化参数迁移到目标领域,实现快递包裹X光图像分类;最后,将相同数据集作为三种模型的输入并对结果进行对比。实验结果表明,TL-ResNet18模型的局部微调和全局微调的识别准确率分别为93.5%、95.0%,相比于ResNet18模型提高了7%、8.5%,且精确度、召回率和F1值都优于ResNet18模型,该方法性能更优,且不受小型数据集对深层网络训练的限制,有利于快递包裹X光图像识别的智能化发展。

MRI与X线诊断交感神经型颈椎病的准确率及影像学特征研究

目的:分析MRI与X线诊断交感神经型颈椎病(SCS)的准确率及影像学特征。方法:选定本院2020年1月至2021年1月接诊的60例高度疑似SCS患者,分别给予X线、MRI检查,将试验检查、肌电图检查等综合临床诊断结果作为本次研究金标准,比较X线、MRI诊断准确率、灵敏度、特异度,Kappa检验X线、MRI与金标准的一致性。绘制ROC曲线,分析X线与MRI诊断效能,比较X线、MRI颈椎曲度异常分型检出率。结果:MRI诊断准确率(91.67%)、灵敏度(92.45%)、特异度(85.71%)均高于X线(58.33%、64.15%、14.29%)(P<0.05)。Kappa检验X线与金标准的一致性一般(Kappa值=0.586),MRI与金标准的一致性较好(Kappa值=0.785)(P<0.05)。X线颈椎曲度反弓(8.33%)、减小(21.67%)、S型(8.33%)检出率与MRI(20.00%、30.00%、20.00%)比较(P>0.05);X线颈椎曲度增大(28.33%)、垂直(33.33%)检出率高于MRI(13.33%、16.67%)(P<0.05)。结论:MRI可提高SCS诊断准确率、灵敏度及特异度,还可提供椎间盘突出或膨出、横突孔狭窄、侧隐窝狭窄、后纵韧带增厚、黄韧带增厚、脊髓水肿或变性等影像学征象,临床价值较高。但X线在颈椎曲度增大、垂直方面的检出率高于MRI,临床医生在具体诊断过程中,应结合患者实际情况针对性地选择诊断方法。

采用X射线三维重构技术检测厚皮柑橘的体积可食率

针对传统无损检测技术无法定量分析厚皮柑橘体积可食率的问题,该研究开发了一套线阵X射线图像采集和三维重构装置,包括果实旋转升降、数据采集、辐射防护及运动控制,实现厚皮柑橘果实的体积可食率检测。以“不知火”柑橘为检测对象,利用X射线投影图的信息熵为评价依据,根据果实大小对检测参数进行优化,得到X射线源的管电压为67 kV,管电流为0.92 mA,线阵探测器的积分时间为1 ms。以旋转角度2.0°为间隔,在圆周方向采集180幅X射线投影图并生成正弦图,通过滤波反向投影算法将正弦图重构为切面图,再利用图像滤波、图像增强、阈值分割对切面图进行图像分割处理,分割出背景区域、果皮区域、果肉区域和空腔区域,然后基于区域面积比定义切面图像可食率。同时,测量了柑橘基本物理参数,计算了切面图像可食率,并建立了与柑橘体积可食率之间的相关关系。结果表明,切面图像可食率与体积可食率相关性最高,相关系数为0.93。最后,选择切面图像可食率作为数学模型的输入特征,利用线性回归模型对厚皮柑橘体积可食率进行定量分析,其预测集决定系数、预测集均方根误差、相对分析误差分别为0.86、4.81%和2.71。综上,利用X射线三维重构技术对厚皮柑橘体积可食率进行无损检测可行,可为其他农产品内部品质无损检测提供参考。

基于改进ResNet50的钨矿石双能X射线图像分选方法

文中提出一种基于深度扩张可分离卷积和注意力机制的残差网络模型(DWAtt-ResNet),通过实验对比表明,该模型在钨矿石双能X射线图像数据集上准确率、F1分数、AUC值和AP值均优于ConvNeXt、DenseNet121和EfficientNet_b4等主流的图像分类模型。通过消融实验表明,该模型准确率达到87.4%,计算量为2.7GFLOPs,参数量为16.95M,相比ResNet50准确率提高3%,计算量降低1.42 GFLOPs,参数量降低6.56M,准确率提升的同时,效率大幅提升,更适合工业生产的矿石快速分拣需求。

用于X射线像增强器便携成像的观瞄目镜设计研究

结合X射线像增强器便携成像这一特殊需求,分析了针对该类观瞄目镜开展人体工程学设计的必要性。从像增强器器件的层面,给出了该类观瞄目镜的系统设计约束与要求。在此基础上,给出了一个该类观瞄目镜的具体设计示例,分析了该类光学系统的具体优化设计方法,并分别从厚度、曲率半径、圆锥系数3个维度开展了具体的系统优化设计。最终,基于光学传递函数评价方法,逐步地对比分析了优化设计与系统性能提升之间的内在联系与对应关系。

基于X射线成像原理的电缆缓冲层缺陷的可视化检测技术对比研究

近年来,高压电缆缓冲层烧蚀缺陷导致电缆故障问题频发,如何有效地对缓冲层缺陷进行检测引起了行业内的广泛关注。基于缓冲层缺陷与其相邻层间的密度差异,分别对X射线数字成像技术和计算机断层成像技术开展了检测参数研究,对比分析了两种方法对不同类型缺陷的检出有效性及在不同干扰下的检测效果。结果表明,计算机断层成像技术在检测准确度和抗干扰性优于X射线数字成像技术,但在检测的现场适用性方面有局限。研究给出两种检测技术的最佳检测参数及各自的优缺点,为后续高压电缆的缓冲层缺陷可视化检测提供了有益的参考。

自适应与多尺度特征融合的X光违禁品检测

针对X光安检违禁品图像空间多尺度变化、背景干扰及模型复杂等问题,提出了空间自适应与多尺度特征融合的YOLOv5轻量模型。以YOLOv5为基本框架,引入自适应空间特征融合机制抑制特征尺度差异的影响,结合双向特征金字塔网络集成了特征双向加权融合;采用轻量化通道注意力机制获得编码的位置信息,增强有效特征的表达;同时利用GhostConv替换部分Conv降低网络计算复杂度。此模型在OPIXray、SIXray、HiXray等3个公开数据集上mAP分别达到94.2%、92.8%、83.3%,比基线模型分别提高了5.4、0.5、1.7个百分点,且未明显改变推理效率,较好兼顾了模型检测精度与速度,优于当前诸多先进算法。