A deep learning fusion model for accurate classification of brain tumours in Magnetic Resonance images

Detecting brain tumours is complex due to the natural variation in their location, shape, and intensity in images. While having accurate detection and segmentation of brain tumours would be beneficial, current methods still need to solve this problem despite the numerous available approaches. Precise analysis of Magnetic Resonance Imaging (MRI) is crucial for detecting, segmenting, and classifying brain tumours in medical diagnostics. Magnetic Resonance Imaging is a vital component in medical diagnosis, and it requires precise, efficient, careful, efficient, and reliable image analysis techniques. The authors developed a Deep Learning (DL) fusion model to classify brain tumours reliably. Deep Learning models require large amounts of training data to achieve good results, so the researchers utilised data augmentation techniques to increase the dataset size for training models. VGG16, ResNet50, and convolutional deep belief networks networks extracted deep features from MRI images. Softmax was used as the classifier, and the training set was supplemented with intentionally created MRI images of brain tumours in addition to the genuine ones. The features of two DL models were combined in the proposed model to generate a fusion model, which significantly increased classification accuracy. An openly accessible dataset from the internet was used to test the model's performance, and the experimental results showed that the proposed fusion model achieved a classification accuracy of 98.98%. Finally, the results were compared with existing methods, and the proposed model outperformed them significantly.

Performance Analysis of Mixed Amplify-and-Forward and Decode-and-Forward Protocol in Underlay Cognitive Networks

In this paper, we propose and evaluate outage performance of a mixed amplify-and-forward(AF) and decode-and-forward(DF) relaying protocol in underlay cognitive radio. Different from the conventional AF and DF protocols, in the proposed protocol, a secondary source attempts to transmit its signal to a secondary destination with help of two secondary relays. One secondary relay always operates in AF mode, while the remaining one always operates in DF mode. Moreover, we also propose a relay selection method, which relies on the decoding status at the DF relay. For performance evaluation and comparison, we derive the exact and approximate closedform expressions of the outage probability for the proposed protocol over Rayleigh fading channel. Finally, we run Monte Carlo simulations to verify the derivations. Results presented that the proposed protocol obtains a diversity order of three and the outage performance of our scheme is between that of the conventional underlay DF protocol and that of the conventional underlay AF protocol.

Effects of sampling frequency on node mobility prediction in dynamic networks: A spectral view

The field of mobility prediction has been widely investigated in the recent past,especially the reduction of the coverage radius of cellular networks,which led to an increase in hand-over events.Changing the cell coverage very frequently,for example,may lead to service disruptions if a predictive approach is not deployed in the system.Although several works examined mobility prediction in the new-generation mobile networks,all of these studies focused on studying the time features of mobility traces,and the spectral content of historical mobility patterns was not considered for prediction purposes as yet.In the present study,we propose a new approach to mobility prediction by analyzing the effects of a proper mobility sampling frequency.The proposed approach lies in the mobility analysis in the frequency domain,to extract hidden features of the mobility process.Thus,we proposed a new methodology to determine the spectral content of mobility traces(considered as signals)and,thus,the appropriate sampling frequency,which can provide numerous advantages.We considered several types of mobility models(e.g.pedestrian,urban,and vehicular),containing important details in the time and frequency domains.Several simulation campaigns were performed to observe and analyze the characteristics of mobility from real traces and to evaluate the effects of sampling frequency on the spectral content.

Natural Rubber Based Composites Comprising Different Types of Carbon-Silica Hybrid Fillers. Comparative Study on Their Electric, Dielectric and Microwave Properties, and Possible Applications

The paper presents a comparative study on the electric, dielectric and microwave properties of natural rubber based composites comprising dual phase fillers prepared from furnace carbon black or conductive carbon black with a different amount of silica. It has been established that, the specifics of the carbon phase have a marked strong effect upon the properties mentioned above. The interpenetration of the two filler phases and the grade of isolation of the conductive carbon phase by the dielectric one depend on the ratio between them. On the other hand, that leads to a change in all properties of the studied composites, which allows tailoring those characteristics.

Hybrid In-Vehicle Background Noise Reduction for Robust Speech Recognition:The Possibilities of Next Generation 5G Data Networks

This pilot study focuses on employment of hybrid LMS-ICA system for in-vehicle background noise reduction.Modern vehicles are nowadays increasingly supporting voice commands,which are one of the pillars of autonomous and SMART vehicles.Robust speaker recognition for context-aware in-vehicle applications is limited to a certain extent by in-vehicle back-ground noise.This article presents the new concept of a hybrid system which is implemented as a virtual instrument.The highly modular concept of the virtual car used in combination with real recordings of various driving scenarios enables effective testing of the investigated methods of in-vehicle background noise reduction.The study also presents a unique concept of an adaptive system using intelligent clusters of distributed next generation 5G data networks,which allows the exchange of interference information and/or optimal hybrid algorithm settings between individual vehicles.On average,the unfiltered voice commands were successfully recognized in 29.34%of all scenarios,while the LMS reached up to 71.81%,and LMS-ICA hybrid improved the performance further to 73.03%.

Analysis and Improvement of the Real-Time Segmented Pulse Compression-Detection Algorithm

Real-Time segmented pulse compression-detection is one of the key technologies of space-borne tracking receiver. Its implementation requires an optimized and dedicated hardware. The real-time processing places several constraints such as area occupied, power comumption, and speed. A number of segmented compression techniques have been proposed to overcome these limitations and decrease the processing latency. However, relatively high power loss in the partial field could limit their implementation in many current real-time systems. A good theoretical model was designed with intersection signal accumulation to enhance signal- noise-ratio （SNR） gain of detecting signal in the paper. From the experimental results it is known that this approach works well for pulse compression-detection, which is better suited for implementation in the high performance of current field programmable gate array （FPGA） with dedicated hardware multipliers.

Comparison of Vertical Handover Decision-Based Techniques in Heterogeneous Networks

Industry leaders are currently setting out standards for 5G networks projected for 2020 or even sooner. Future generation networks will be heterogeneous in nature as no single network type will be capable of optimally meeting all the rapid changes in customer demands. With the advent of multi-homed devices and heterogeneous network (HetNet) solution, great concerns arise in the processes involved for successful handover. Active calls that get dropped or cases of poor quality of service experienced by mobile users can be attributed to the phenomenon of delayed handover (HO) or an outright case of an unsuccessful handover procedure. This work compares multiple criteria handover basis to its traditional single relative signal strength (RSS) base counterpart. It analyses the performance of a fuzzy-based VHO algorithm scheme in a Wi-Fi, WiMAX, UMTS and LTE integrated network using OMNeT++ event simulator. The loose coupling network architecture is adopted and simulation results analysed for the two major categories of handover;the multiple and single criteria. Results obtained show a better overall throughput, better call dropped rate and shorter handover time for the multiple criteria based decision method as compared to the single criteria based technique. This work also highlights current research trends, challenges of seamless handover and initiatives for Next Generation HetNet.

Highly Efficient Conversion of Solar Energy by the Photoelectric Converter and a Thermoelectric Converter

The analysis of the principle operation of the solar element is given in the work. The efficiency of the creation of combined high-efficiency converters of network energy in electric and thermal is indicated. A method for creating a combined photo thermo converter with a high value of the efficiency of a solar element has been found. A method for separating light from optical lenses is shown. The calculation of the light intensity into diffraction pattern is calculated using the Huygens-Fresnel principle. A new design of a highly efficient combined light-thermal converter into an electrical, with a solar element, operating on selective photoactive radiation is presented. The process of conversion of non-active radiation into electrical radiation by means of a thermo electronic converter is described. The ways of the solution of the problem connected with the reduction of the coefficient of its full effect as a function of temperature and characteristics are indicated.

The Effect of Speech Fragmentation and Audio Encodings on Automatic Parkinson’s Disease Recognition

Parkinson’s disease is a neurological disease which is incurable according to current clinical knowledge. Therefore, early detection and provision of appropriate treatment are of primary importance. Speech is one of the biomarkers that enable the detection of Parkinson’s disease affection. Numerous researches are based on recordings from controlled environments;nonetheless fewer apply real circumstances. In the present study, three objectives were examined: recording fragmentation (paragraph, sentences, time-based), variable encodings (Pulse-Code Modulation [PCM], GSM-Full Rate [FR], G.723.1) and majority voting on 8 kHz records using multiple classifiers. Support Vector Machine (SVM), Long Short-Term Memory (LSTM), i-vector and x-vector classifiers were evaluated in contrast with SVM as baseline. The highest results in accuracy and F1-score were achieved using i-vector models. Although variable encodings generally caused decrease in Parkinson-disease recognition, decline was within 2% - 3% at best. Moreover, fragmentation did not yield a clear outcome though some classifiers performed with the very similar efficiency along the differently fragmented sets. Majority voting did produce a slight increase in classification performance compared to as if no aggregation is used.

When Mulan meets Romeo： Cultural impact upon the dating scripts, perceptions and behaviors of college students from the United States and Taiwan in their cross-cultural romantic relationships

This study examines the cultural impact upon the dating scripts, perceptions, and behaviors of college students from the United States and Taiwan in their cross-cultural romantic relationships. Ten same-culture and ten cross-culture couples participated in the retrospective account interviews. Reconfirming the prevalent quantitative research findings that the American participants tend to begin dating at a younger age and change dating partners more frequently, this qualitative study finds that most of the American and Taiwan Residents participants perceived their dating and cross-cultural romantic relationships seriously by demonstrating a strong sense of responsibility and consistent support in the face of language difficulties, cultural confusions, and social obstacles.

Complex Permittivity and Permeability Studies Viewing Antenna Applications of NBR-Based Composites Comprising Conductive Fillers

The work presents studies on the complex permittivity and permeability of composites based on acrylonitrile butadiene rubber containing combinations of conductive fillers which include carbon black and nickel powder. The properties of those composites, containing each of the fillers at the same amount were compared. The permittivity and permeability values of the composites are influenced remarkably by their morphology and structure as well as by the morphological and structural specifics of both fillers. As electron scanning microscopy studies confirm, those parameters are predetermined by the nature of the composites studied—particle size, particles arrangement in the matrix and their tendency to clustering. Last but not least matrix-filler interface phenomena also impact the characteristics in question. The possibilities for applications of the composites in antennae have been studied, in particular, as substrates and insulating layers in flexible antennae for body centric communications (BCCs). The research results allow the conclusion that these materials can find such applications indeed. Composites of higher conductivity can be used where surface waves are generated to provide on-body communications, while composites of lower conductivity may be used for antennae that will be on the body of a person and will transmit to and receive from other antennas that are not on the body of the same person (off-body communications). It is clear that one can engineer the properties of antennae substrates at microwave frequencies by adjusting the filler content and the type of filler and thus control and tailor the antenna performance specific for a particular application.

Optimization of Quantizer’s Segment Threshold Using Spline Approximations for Optimal Compressor Function

In this paper, the optimization of quantizer’s segment threshold is done. The quantizer is designed on the basis of approximative spline functions. Coefficients on which we form approximative spline functions are calculated by minimization mean square error (MSE). For coefficients determined in this way, spline functions by which optimal compressor function is approximated are obtained. For the quantizer designed on the basis of approximative spline functions, segment threshold is numerically determined depending on maximal value of the signal to quantization noise ratio (SQNR). Thus, quantizer with optimized segment threshold is achieved. It is shown that by quantizer model designed in this way and proposed in this paper, the SQNR that is very close to SQNR of nonlinear optimal companding quantizer is achieved.

Correlation between Electrical Conductivity and Microwave Shielding Effectiveness of Natural Rubber Based Composites, Containing Different Hybrid Fillers Obtained by Impregnation Technology

The paper presents the synthesis and characterization of carbon black/silicone dioxide hybrid fillers obtained by an impregnation technology. The electromagnetic interference shielding effectiveness of the composites filled with carbon black/silicone dioxide hybrid fillers was measured in wide frequency range of 1 - 12 GHz. The dc and ac electrical conductivity of composites also have been investigated. The relationship between electrical (dc and ac) conductivity and shielding effectiveness was analyzed. A positive correlation was found between the absorptive shielding effectiveness and ac conductivity for composites comprising conductive carbon black/silica filler, when the filler loading is above the percolation threshold.

A New Perspective on Time and Gravity

This paper presents a hypothesis regarding the existence of time fused in spacetime, assuming that time possesses the properties of both a particle and a field. This duality is referred to as the field-particle of time (FPT). The analysis shows that when the FPT moves through matter, it causes time dilation. The FPT is also a significant element that appears in relativistic kinetic energy (KE = (γ - 1) · mc2). Accelerating matter to near the speed of light requires relativistic energy approaching infinity, which corresponds to the relativistic kinetic energy. Meanwhile, the potential energy (PE = mc2) from the rest mass remains constant. Then, the mass-energy equation can be rearranged in terms of PE and KE, as shown in E = (1 + (γ - 1)) · mc2. The relativistic energy of the FPT also directly affects the gravitational attraction of matter. It transfers energy to each other through spacetime. The analysis demonstrates that the gravitational force is inversely proportional to the distance squared, following Newton’s law of gravity, and it varies with the relative velocity of matter. The relationship equation between relative time and the gravitational constant indicates that a higher intensity of the gravitational field leads to a slower reference time for matter, in accordance with the general theory of relativity. A thought experiment presents a comparison of two atomic clocks placed in different locations. The first one is placed in a room temperature, around 25°C, on the surface of the Earth, and the second one is placed in high-density areas. The analysis, considering the presence of the FPT, shows that the reference time slows down in high-density areas. Therefore, the second clock must be noticeably slower than the first one, indicating the existence of the FPT passing through both atomic clocks at different speeds.

数字图像识别在混合油类三维荧光光谱分析中的应用

海上溢油已成为全球环境污染的重要问题之一,溢油严重破坏了海洋生态的平衡,并导致人类健康受到危害。因此,研究高效的溢油检测方法对保护海洋生态环境具有重要意义。三维荧光光谱技术因能获得溢油的“指纹”图谱而成为溢油鉴别领域的有效分析手段,其与平行因子分析算法相结合获得了良好的溢油鉴别效果。但平行因子算法在使用过程中需要确定不同石油产品本身所适用的浓度范围,且其对预估计组分数敏感,组分数选择是否准确直接影响最终定性定量结果,这些问题都会对油类检测造成使用上的限制。油类组分极为复杂,其中各组分间不存在统一的线性浓度范围,其相互之间还受到荧光猝灭效应的影响。直接对未经稀释的油类样本进行光谱数据采集,所获得的三维荧光光谱会因样本中组分的种类及其含量不同而存在较大差异,导致对三维荧光光谱数据进行解析的平行因子分析算法不再适用。但组分的种类及含量相近的油样其光谱特征相似度较高,并且随着特定组分及其含量的改变,其光谱形状的变化规律也较为明显。基于此,将三维荧光光谱和数字图像识别相结合,提出一种针对混合油类样本的辨识方法。首先,利用五种矿物油(汽油、柴油、航空煤油、机油和润滑油)配制三类混合油样本,其中每类混合油是用其中两种不同矿物油以不同体积比直接混合配制而成;然后利用FS920荧光光谱仪获取样本的三维荧光光谱数据,并对该数据进行求导及灰度化预处理,进而得到三维荧光导数光谱灰度图;其次提取样本三维荧光导数光谱灰度图的颜色、纹理和形状等数字图像特征;最后,通过Fisher判别分析建立样本的分类模型,采用逐步回归建立混合油样本各组分相对体积的定量模型。分类模型对三类混合油样本的分类及识别效果良好。所建立的定量模型的线性相关性R大于0.99,显著性检验p值小于0.05。研究结果表明,三维荧光光谱的数字图像特征可以被本文所述方法有效提取并用于对油类样本的定性定量分析。该研究为海面溢油检测提供了一种简单、可靠的识别方法。

海面溢油三维荧光光谱消除瑞利散射方法的研究

三维荧光光谱分析法以其灵敏度高、选择性好、操作简单和可用于多组分混合物分析等优点成为诸多研究者在海面溢油鉴别中的热点选择。但三维荧光光谱中存在的瑞利散射会对光谱的准确检测产生较大的影响,因此有效地消除瑞利散射对后续光谱的定性鉴别和定量分析具有重要意义。采用仪器校正法、空白扣除法、 Delaunay三角形内插值法和缺损数据重构(MDR)法对海面溢油三维荧光光谱中的瑞利散射进行校正。首先以海水的SDS胶束溶液作为溶剂,将航空煤油和润滑油按不同相对体积分数比配制8个校正样本和3个测试样本;然后利用FS920稳态荧光光谱仪采集11个样本的三维荧光光谱数据,并分别采用仪器校正法、空白扣除法、 Delaunay三角形内插值法和缺损数据重构(MDR)法消除瑞利散射的干扰;再利用核一致诊断法估计出最佳的组分数;最后利用平行因子分析(PARAFAC)对混合油样本的三维荧光光谱数据进行定性鉴别和定量分析。研究结果表明:采用发射波长滞后激发波长以消除瑞利散射的仪器校正法会丢失部分有效光谱信息;采用空白扣除法无法彻底消除瑞利散射,在光谱中仍然存在散射干扰,利用PARAFAC解析后得到的激发、发射光谱会出现失真,且预测的浓度值偏差较大;采用Delaunay三角形内插值法消除瑞利散射后,利用PARAFAC解析所得到的激发、发射光谱与真实光谱吻合度较高,且预测的浓度值偏差较小;而采用MDR消除瑞利散射后,利用PARAFAC解析所获得的激发、发射光谱与真实光谱吻合度最高,且相较于其他几种方法预测的浓度值偏差最小,得到的样本回收率为98.9%和100%,预测均方根误差均小于等于0.130。根据定性鉴别、定量分析的结果, MDR能够在保证原有特征光谱不失真的基础上有效消除瑞利散射带来的影响,是一种消除三维荧光光谱数据中瑞利散射较为理想的方法。

结合平行因子分析算法和模式识别方法的三维荧光光谱技术用于石油类污染物的检测

随着海洋中石油资源的不断开发,泄漏到海洋环境中的石油也日益增多,它不仅威胁着海洋生态环境,同时也严重影响着人们的身体健康。因此,快速、有效地检测出海洋环境中的石油类污染物对于保护海洋生态环境和人类健康具有重要意义。石油产品中含有大量的多环芳烃,其具有较强的荧光特性。因此,荧光光谱技术成为检测石油类污染物的重要手段之一。利用三维荧光光谱技术结合平行因子分析算法和模式识别方法,对石油类污染物进行表征和分类。首先,以海水和十二烷基硫酸钠(SDS)配制的胶束溶液作为溶剂,分别配制不同浓度的柴油、航空煤油、汽油和润滑油溶液,最终得到80个实验样本;然后,利用FLS920型荧光光谱仪采集实验样本的三维荧光光谱数据,并通过Delaunay三角形内插值法对所获得的三维荧光光谱数据进行去散射处理;其次,利用平行因子分析(PARAFAC)算法分解去散射后的三维荧光光谱数据,通过运用核一致诊断法和残差分析法对组分数进行估计;最后,为了建立稳健的分类模型,利用Kennard-Stone算法将80个实验样本分为60个训练集样本和20个测试集样本,运用K最近邻(KNN)算法、主成分判别分析(PCA-LDA)算法以及偏最小二乘判别分析(PLS-DA)算法分别建立分类模型,并利用灵敏度、特异性和准确率对分类效果进行评估。研究结果表明:三种分类模型对测试集中样本的识别准确率分别为85%, 90%和94%,其中, PLS-DA分类模型对测试集样本的识别准确率最高,具有最佳的分类效果。因此,在利用平行因子分析算法提取石油类污染物荧光光谱数据的基础上,结合模式识别方法可以很好的对不同种类油品进行分类研究。利用三维荧光光谱技术结合平行因子分析算法和模式识别方法快速、有效地检测油类污染物,为石油类污染物的快速检测提供了一种新的研究思路和重要参考。

基于三维荧光光谱技术结合交替加权残差约束四线性分解的不同盐度条件下混合油液检测

石油作为一种重要的化石能源,是人类社会生产活动中不可缺少的一部分。石油在被人们开采、使用的过程中不可避免地会发生泄漏,泄漏的石油会给生态环境带来严重的威胁。因此,在石油泄漏后需要及时对其进行处理,而其前提是能够准确识别石油种类。由于石油中多种物质具有荧光特性,因此应用荧光光谱法可对石油进行有效检测。但石油所含组分较多,使得其光谱信息重叠严重,识别困难。而三阶校正方法具有"三阶优势",可以分辨高共线性、高噪声水平下的数据。其中,三阶校正中的交替加权残差约束四线性分解(AWRCQLD)算法具有收敛速度快、对组分数不敏感等优点;因此,利用三维荧光光谱技术结合AWRCQLD算法,对混合油液进行检测。首先,配制3种盐度条件下的十二烷基硫酸钠(SDS)溶剂;并在每种盐度条件下分别将航空煤油和润滑油按照不同浓度比混合,最终得到24个校正样本和9个预测样本。然后,使用FLS920荧光光谱仪对实验样本进行光谱数据采集。其次,使用扣除空白法去除光谱中的散射,并通过核一致诊断法判断混合油中的组分数。最后,用AWRCQLD算法对四维光谱矩阵进行解析。研究结果表明,在0~20盐度范围内,随着盐度的增加,航空煤油的荧光强度先减小后增大,润滑油的荧光强度先增大后减小;混合油解析光谱曲线分别与航空煤油及润滑油的实际光谱曲线重合度良好;经AWRCQLD算法解析后得到的航空煤油的回收率范围为100.2%~109%,均方根误差为0.0021 mg·mL^-1;润滑油的回收率范围为91.8%~109.3%,均方根误差为0.0048 mg·mL^-1。通过引入盐度作为新一维度的数据,从而将三维光谱数据阵扩展到相应的四维光谱数据阵。并利用AWRCQLD算法对四维光谱数据阵进行了解析,实现了在不同盐度条件下对混合油的定性和定量分析。同时,为不同盐度条件下的混合油液检测提供了参考。

通信网可靠性的内涵及其测度方法

通信网可靠性的定义是我们开展通信网可靠性研究的基础。本文明确了通信网可靠性的研究目的和内容,分析了通信网可靠性定义的六个要素,并给出了通信网可靠性的完整定义。最后介绍了通信网可靠性测度的方法。

基于重构三维荧光光谱结合偏最小二乘判别分析的油类识别方法研究

油类污染日渐频繁,给人类健康及生态环境造成了严重的威胁。因此,研究有效的油类识别方法对保护生态环境具有重要意义。三维荧光光谱技术是识别油类最有效的分析手段之一,利用二阶校正方法对三维荧光光谱数据进行解析,然后利用模式识别对二阶校正方法解析结果中的浓度得分矩阵进行分类,可以实现对未知样本的定性识别。然而,此类方法在对未知样本进行分类识别的过程中,只应用了浓度得分矩阵,其本质上只是利用样本所含化学成分的相对含量差异对未知样本进行了分类。并没有利用具有定性意义的载荷矩阵,即没有从样本所含化学成分本身实现对样本的定性。基于此,将重构的三维荧光光谱和偏最小二乘判别分析(PLS-DA)相结合,提出了一种针对油类样本的辨识方法。首先,利用四种油类(汽油、柴油、航空煤油和润滑油)在不同的背景环境下(纯净水、自来水、河水及海水配制的十二烷基硫酸钠溶剂)配制了80个油类样本;然后,利用FS920荧光光谱仪采集样本的三维荧光光谱数据,并对该数据进行去散射及标准化预处理;其次,利用Leverage值识别并删除其中的异常光谱,并利用平行因子分析算法(PARAFAC)对剩余的光谱进行重构;最后,通过PLS-DA建立重构三维荧光光谱的分类模型;并将重构与未重构的三维荧光光谱分别建立的分类模型进行了对比。分析结果表明,三维荧光光谱经过重构后,可以将四种油类的正确分类率分别从原来的100%, 50%, 60%和20%提高到100%, 100%, 100%和100%,表明重构的三维荧光光谱具有更加明显的类内特征。重构三维荧光光谱所建立的分类模型的灵敏度(SENS)、特异性(SPEC)及F分数分别为100%, 100%和100%,表明所建立的模型具有稳健及可靠的分析结果。该研究中,重构三维荧光光谱利用了PARAFAC解析结果中的浓度得分矩阵及载荷矩阵,所建立的PLS-DA分类模型不仅从化学成分相对含量的差异而且从化学成分本身对样本进行了定性识别,所得结果更加具有说服力。该研究为油类识别提供了一种可靠的方法。