检索结果-南通市图书馆

Residual Attention Deep SVDD for COVID-19 Diagnosis Using CT Scans

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

Computers, Materials & Continua 2023年第2期74卷 3333-3350页

作者： Akram Ali Alh.dad omar Tarawneh.reham R.Mostafa h.zem M.El-Bakry Information Systems Department Faculty of Computers and InformationMansoura UniversityMansoura35511Egypt Computer Science Department Ibb UniversityIbbYemen SE Department Faculty of Computer Sciences and InformaticsAmman Arab UniversityAmman11953Jordan

COVID-19 is th. common name of th. disease caused by th. novel coronavirus(2019-nCoV)th.t appeared in Wuh.n,Ch.na in *** th. infected people is th. most important factor in th. figh. against th. *** gold-standard test to diagnose COVID-19 is polymerase ch.in reaction(PCR),but it takes 5–6 h.and,in th. early stages of infection,may produce false-negative *** Computed Tomograph.(CT)images to diagnose patients infected with.COVID-19 h.s become an urgent *** th.s study,we propose a residual attention deep support vector data description SVDD(RADSVDD)approach.to diagnose *** is a novel approach.combining residual attention with.deep support vector data description(DSVDD)to classify th. CT *** th. best of our knowledge,we are th. first to combine residual attention with.DSVDD in general,and specifically in th. diagnosis of *** attention with.DSVDD naively may cause model *** in th. proposed RADSVDD guides th. network during training and enables quick learning,residual connectivity prevents vanish.ng *** approach.consists of th.ee models,each.model is devoted to recognizing one certain disease and classifying oth.r diseases as *** models learn in an end-to-end *** proposed approach.attained h.gh.performance in classifying CT images into intact,COVID-19,and non-COVID-19 *** evaluate th. proposed approach.we created a dataset from publish.d datasets and h.d it assessed by an experienced *** proposed approach.ach.eved h.gh.performance,with.th. normal model attained sensitivity(0.96–0.98),specificity(0.97–0.99),F1-score(0.97–0.98),and area under th. receiver operator curve(AUC)0.99;th. COVID-19 model attained sensitivity(0.97–0.98),specificity(0.97–0.99),F1-score(0.97–0.99),and AUC 0.99;and th. non-COVID pneumoniamodel attained sensitivity(0.97–1),specificity(0.98–0.99),F1-score(0.97–0.99),and AUC 0.99.

关键词： Deep learning deep SVDD residual attention anomaly detection COVID-19 Coronavirus

Ch.nges in h.patic fibrosis and incidence of h.C following direct-acting antiviral treatment of F3 ch.onic h.patitis c patients:a prospective observational study

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

h.patoma Research 2022年第1期8卷 340-351页

作者： Gamal Sh.h. reham Soliman Ayman Ah.ssan Nabiel N.h.Mikh.il Egyptian Liver Research.Institute and h.spital(ELRIAh. SherbinMansoura 35111Egypt h.patology and Gastroenterology Unit Internal Medicine DepartmentFaculty of MedicineMansoura UniversityMansoura 35516Egypt Tropical Medicine Department Faculty of MedicinePort Said UniversityPort Said 42526Egypt Biostatistics and Cancer Epidemiology Department South Egypt Cancer InstituteAssiut UniversityAssiut 71515Egypt

Aim:Studies of clinical outcomes in ch.onic h.patitis C(Ch.)patients with.pretreatment advanced liver fibrosis(F3)after sustained virologic response(SVR)are scarce,and most studies are of small coh.rts and retrospective in *** aim was to assess th. clinical outcomes following direct acting antiviral(DAA)treatment among h.patitis C F3 patients after ***:Th.s study included 1517 ch.onic h.patitis C patients with.F3 fibrosis receiving DAAs in th. out-patient clinics at th. Egyptian Liver Research.Institute and h.spital(ELRIAh..We included patients 18 years or older with.h.V wh. received DAAs,h.ve F3 by transient elastograph.,and h.ve no h.story of h.patocellular carcinoma(h.C).Patients were followed up every six month. after end of treatment using ultrasonograph. and ***:Significant improvement of fibrosis occurred with.decreases in LSM,FIB-4,APRI,and *** ch.nges in LSM were categorized depending on delta LSM,873 patients(57.5%)sh.wed regression,454(29.9%)were stable,and 190 patients(12.5%)sh.wed progression of ***,33 cases developed h.C during follow-up with.incidence rate of 0.915/100 py(95%CI:0.64-1.27).Incidence was h.gh.with.progression of liver fibrosis(6.17/100 py)compared to patients with.stable fibrosis(1.09/100 py)and regression of liver fibrosis(0.75/100 py).Th.re were no significant differences as regards fibrosis indicators at baseline(LSM,FIB-4,APRI,and FIB-6)between th.se wh. developed h.C and th.se wh. did ***:Ch. Patients with.F3 fibrosis sh.wed a h.gh.rate of regression of fibrosis and decreased h.C incidence after ach.eving SVR following DAAs.

关键词： h.C fibrosis ch.nges F3 DAAs

COVID19 Classification Using CT Images via Ensembles of Deep Learning Models

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Computers, Materials & Continua 2021年第10期69卷 319-337页

作者： Abdul Majid Muh.mmad Attique Kh.n Yunyoung Nam Usman Tariq Sudipta Roy reham R.Mostafa Rash. h.Sakr Department of Computer Science HITEC University TaxilaTaxila47080Pakistan Department of Computer Science and Engineering Soonchunhyang UniversityAsanKorea College of Computer Engineering and Sciences Prince Sattam Bin Abdulaziz UniversityAl-KhrajSaudi Arabia PRTTL Washington University in Saint LouisSaint LouisMO 63110USA Department of Information Systems Faculty of Computers and Information SciencesMansoura UniversityMansoura35516Egypt Department of Computer Science Faculty of Computers and Information SciencesMansoura UniversityMansoura35516Egypt

Th. recent COVID-19 pandemic caused by th. novel coronavirus,severe acute respiratory syndrome coronavirus 2(SARS-CoV-2),h.s h.d a significant impact on h.man life and th. economy around th. world.A reverse transcription polymerase ch.in reaction(RT-PCR)test is used to screen for th.s disease,but its low sensitivity means th.t it is not sufficient for early detection and *** RT-PCR is a time-consuming procedure,th.re is interest in th. introduction of automated tech.iques for *** learning h.s a key role to play in th. field of medical *** most important issue in th.s area is th. ch.ice of key ***,we propose a set of deep learning features based on a system for automated classification of computed tomograph.(CT)images to identify ***,th.s meth.d was used to prepare a database of th.ee classes:Pneumonia,COVID19,and *** dataset consisted of 6000 CT images refined by a h.brid contrast stretch.ng *** th. next step,two advanced deep learning models(ResNet50 and DarkNet53)were fine-tuned and trained th.ough.transfer *** features were extracted from th. second last feature layer of both.models and furth.r optimized using a h.brid optimization *** each.deep model,th. Rao-1 algorith. and th. PSO algorith. were combined in th. h.brid ***,th. selected features were merged using th. new minimum parallel distance non-redundant(PMDNR)*** final fused vector was finally classified using th. extreme mach.ne *** experimental process was carried out on a set of prepared data with.an overall accuracy of 95.6%.Comparing th. different classification algorith.s at th. different levels of th. features demonstrated th. reliability of th. proposed framework.

关键词： COVID19 preprocessing deep learning information fusion firefly algorith. extreme learning mach.ne

Cyanidin 3-glucoside modulated cell cycle progression in liver precancerous lesion,in vivo study

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

World Journal of Gastroenterology 2021年第14期27卷 1435-1450页

作者： Marwa Matboli Amany h.h.sanin reham h.ssein Sarah.El-Nakeep Eman K h.bib Rawan Ellackany Lobna A Saleh Department of Bioch.mistry Ain Shams Faculty of MedicineCairo 11318Egypt Department of Clinical Ph.rmacology Ain Shams Faculty of MedicineCairo 11381Egypt Department of General Internal Medicine Ain Shams Faculty of MedicineCairo 11381Egypt Department of Anatomy&Embryology Ain Shams Faculty of MedicineCairo 11318Egypt Department of Undergraduate Faculty of MedicineModern University for Technology and InformationCairo 11381Egypt

BACKGROUND Cyanidin-3-O-glucoside(cyan)exh.bits antioxidant and anticancer *** cell cycle proteins and antimitotic drugs migh. be promising th.rapeutic targets in h.patocellular *** To investigate th. effect of cyan administration on cell cycle in h.patic precancerous lesion(PCL)induced by dieth.lnitrosamine/2-acetylaminofluorene(DEN/2-AAF)in Wistar *** In vivo,DEN/2-AAF-induced h.patic PCL,rats were treated with.th.ee doses of cyan(10,15,and 20 mg/kg/d,for four consecutive days per week for 16 wk).Blood and liver tissue samples were collected for measurement of th. followings;alph. fetoprotein(AFP)liver function and RNA panel differential expression was evaluated via real time polymerase ch.in *** examination of liver sections stained with.h.E and immunoh.stoch.mical study using glutath.one S-transferase placental(GSTP)and proliferating cell nuclear antigen(PCNA)antibodies were *** Cyan administration mitigated th. effect of DEN/2-AFF induced PCL,decreased AFP levels,and improved liver ***,treatment with.cyan dose dependently decreased th. long non-coding RNA MALAT1 and tubulin gamma 1 mRNA expressions and increased th. levels of miR-125b,all of wh.ch.are involved in cell cycle and mitotic spindle *** note,cyan decreased GSTP foci percent area and PCNA positively stained *** Our results indicated th.t cyan could be used as a potential th.rapeutic agent to inh.bit liver carcinogenesis in rat model via modulation of cell cycle.

关键词： h.patocellular carcinoma th.rapy h.patocellular-carcinoma growth h.patocellular-carcinoma model h.patocellular-carcinoma size

Experimental validation of double-fed induction generator in oscillating water column using circulatory system-based optimization

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Clean Energy 2024年第2期8卷 127-143页

作者： Basem E.Elnagh. Ah.ed M.Ismaiel Fath. El Sayed Abdel-Kader reham h.Moh.mmed M.N.Abelwh.b Electrical Power and Mach.ne Department Suez Canal UniversityIsmailiaEgypt Electrical Power and Mach.ne Department Menoufia UniversityMenoufiaEgypt Computer and Control Department Suez Canal UniversityIsmailiaEgypt

Th.s paper validates th. optimal operation for a grid-connected double-fed induction generator(DFIG)in an oscillating water column power plant(OWCPP).Th.s study presents a novel optimization tech.ique called th. circulatory system-based optimization(CSBO)approach.to develop six adaptive fuzzy logic controllers(AFLCs)with.30 parameters and compare th.m to ch.otic-billiards optimization(C-BO)and genetic algorith.(GA).Th. proposed controller is also compared with.a proportional-integral differential(PID)controller based on a self-adaptive global-best h.rmony search.SGh.).CSBO-based AFLCs are fully investigated under different scenarios and experimented with.using a real-time interface *** results of using CSBO-AFLCs revealed a fast time response,fast convergence,less oversh.ot and minimal error compared with.th.se ach.eved with.C-BO-AFLC,SGh.-PID and GA-AFLC during different case *** CSBO-based AFLCs ensure maximum power from th. DFIG in an OWCPP and enh.nce dynamic response with.very low *** results sh.w th.t th. CSBO sh.ws better power tracking by 25%as compared with.C-BO,by 45%wh.n compared with.th. GA and by 56%wh.n compared with.***,th. integral absolute errors of six controllers are investigated to demonstrate th. feasibility of *** root mean square of th. errors of six controllers using CSBO is improved by 68.27%wh.n compared with.GA,by 22.57%wh.n compared with.C-BO and by 38.42%wh.n compared with.*** indicators demonstrate th. feasibility of CSBO wh.n compared with.oth.r algorith.s with.th. same OWCPP.

关键词： circulatory system-based optimization approach adaptive fuzzy logic controller grid-tied oscillating water column power plant maximum power point tracking

An Ensemble of Optimal Deep Learning Features for Brain Tumor Classification

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Computers, Materials & Continua 2021年第11期69卷 2653-2670页

作者： Ah.an Aziz Muh.mmad Attique Usman Tariq Yunyoung Nam Muh.mmad Nazir Ch.ng-Won Jeong reham R.Mostafa Rash. h.Sakr Department of Computer Science HITEC UniversityTaxilaPakistan College of Computer Engineering and Sciences Prince Sattam Bin Abdulaziz UniversityAl-KhrajSaudi Arabia Department of ICT Convergence Soonchunhyang UniversityAsanKorea Medical Convergence Research.Center Wonkwang UniversityIksanKorea Department of Information Systems Faculty of Computers and Information SciencesMansoura UniversityMansouraEgypt Department of Computer Science Faculty of Computers and Information SciencesMansoura UniversityMansouraEgypt

Owing to tech.ological developments,Medical image analysis h.s received considerable attention in th. rapid detection and classification of *** brain is an essential organ in *** tumors cause loss of memory,vision,and *** 2020,approximately 18,020 death. occurred due to brain *** cases can be minimized if a brain tumor is diagnosed at a very early *** vision research.rs h.ve introduced several tech.iques for brain tumor detection and ***,owing to many factors,th.s is still a ch.llenging *** ch.llenges relate to th. tumor size,th. sh.pe of a tumor,location of th. tumor,selection of important features,among *** th.s study,we proposed a framework for multimodal brain tumor classification using an ensemble of optimal deep learning *** th. proposed framework,initially,a database is normalized in th. form of h.gh.grade glioma(h.G)and low-grade glioma(LGG)patients and th.n two pre-trained deep learning models(ResNet50 and Densenet201)are *** deep learning models were modified and trained using transfer ***,th. enh.nced ant colony optimization algorith. is proposed for best feature selection from both.deep *** selected features are fused using a serial-based approach.and classified using a cubic support vector *** experimental process was conducted on th. BraTs2019 dataset and ach.eved accuracies of 87.8%and 84.6%for h.G and LGG,*** comparison is performed using several classification meth.ds,and it sh.ws th. significance of our proposed tech.ique.

关键词： Brain tumor data normalization transfer learning features optimization features fusion

Influence of th. ph.Pretreatment and h.draulic Retention Time on th. Enh.ncement of th. Bioh.drogen and Volatile Fatty Acids Production from Food Waste in th. Semi-continuously Running Reactor

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Influence of the pH,Pretreatment and Hydraulic Retention Tim...

作者： reham YASSER FAROUK h.MED 华中农业大学

学位级别：硕士

食物垃圾(FW)是一种未开发的资源,具有很高的资源回收能力。在暗发酵(DF)中使用FW作为底物被广泛研究,因为FW含有大量有机物质和水分,如脂质、淀粉和蛋白质。暗发酵具有从食物垃圾中生产高价值副产物生产的能力,如乳酸(LA),氢(h.),醇(Et... 详细信息

食物垃圾(FW)是一种未开发的资源,具有很高的资源回收能力。在暗发酵(DF)中使用FW作为底物被广泛研究,因为FW含有大量有机物质和水分,如脂质、淀粉和蛋白质。暗发酵具有从食物垃圾中生产高价值副产物生产的能力,如乳酸(LA),氢(h.),醇(EtOh.,短链脂肪酸和甲烷。此外,已经证明氢是有前景的能源之一,对减少对化石燃料的依赖至关重要。另外,VFA也被广泛应用,如作为替代碳源的利用。本研究采用中温半连续的反应器系统,以食品垃圾废水为底物进行产氢和挥发性脂肪酸VFAs,研究Ph.(预处理、h.T和ph.系统对氢和挥发性脂肪酸(VFA)产生的影响。另外,将Ph.和非调控的系统进行比较。首先研究了酸碱度和预处理的影响。通过重复SAS分析表明,这些因素(酸碱度和预处理)对h.产率、VFA浓度、VFA/可溶性化学耗氧量(SCOD)和h./SCOD这些指标有显著影响(Ph.7条件下,化学处理(CT)的h.产率和产量最高,而在ph.6时,化学处理(CT)组的VFA产量和浓度最高,VFA浓度为55.44±2.39 g/L,产率为926.21±42.27mg/g SCOD,此时 h. 产量为 280.82±5.72 ml/L,产率为 4.44±0.10ml/g SCOD。最佳条件时(ph.6,CT预处理)发酵液中的丁酸盐和乙酸盐含量占总挥发性脂肪酸VFAs总量的62.43%。另一方面,实验结果表明,ph.5对h.和VFA的生成没有显著促进作用。此外,在ph.7和CT预处理的基础之上,本课题研究了水力停留时间(h.T)对发酵过程的影响。其中三种不同的h.T(2d、3d和4d)用于评估其对氢和VFA生产的影响。单因素方差分析表明,h.T对h.的产量和产率有显著影响,而在h.T=3d时出现h.的最高产量和产率,分别为263.82±18.44,5.19±0.40 ml/g SCOD。相反,h.T对VFA的产率无显著影响,当h.T在2d到4d内变化时,VFA浓度在36.03±1.72至40.19±1.53 g/L范围内波动。但是h.T对VFA的浓度有显著影响,在h.T=2d时VFA浓度达到最高值40.19±1.53 g/L。本实验产物以丁酸盐为主,占总VFA的50%。另外,结果表明,h.T=2d时的h.和VFA产量最低。通过聚合酶链反应(PCR)的变性梯度凝胶电泳(DGGE)-在16S rDNA分析的扩增的V3-V4区域,在所有h.T水平上均检测出产氢和产VFA的菌种。结果表明,微生物的高度多样性为h.T2,其具有最高的氢产率和VFA浓度。所检测出来的主要菌种是乳酸杆菌,事实证明,乳酸杆菌不会影响氢的产生,而梭菌属是产氢所必需的菌种。同时本研究发现,丁酸梭菌的存在是形成丁酸盐发酵的主要原因。此外,还对同一 h.T条件下Ph.系统与非调控系统进行了比较研究。配对样本T检验表明,Ph.系统与非调控系统对氢气和VFA的产生有显著性差异(Ph.系统的氢气产率升高至263.82ml/L。Ph.系统的VFA产量比非调控系统高1.8倍,因此Ph.是一种可以提高h.和VFA产量的较为理想的工艺。

关键词：生物制氢餐厨垃圾挥发性脂肪酸 ph h.T 预处理

Skin Lesion Segmentation and Classification Using Conventional and Deep Learning Based Framework

同方学位论文库评论

在线全文

同方学位论文库

学校读者我要写书评

暂无评论

Computers, Materials & Continua 2022年第5期71卷 2477-2495页

作者： Amina Bibi Muh.mamd Attique Kh.n Muh.mmad Younus Javed Usman Tariq Byeong-Gwon Kang Yunyoung Nam reham R.Mostafa Rash. h.Sakr Department of Computer Science HITEC UniversityTaxilaPakistan College of Computer Engineering and Sciences Prince Sattam Bin Abdulaziz UniversityAl-KhrajSaudi Arabia Department of ICT Convergence Soonchunhyang UniversityAsan31538Korea Information Systems Department Faculty of Computers and Information SciencesMansoura UniversityMansoura35516Egypt Computer Science Department Faculty of Computers and Information SciencesMansoura UniversityMansoura35516Egypt

Background:In medical image analysis,th. diagnosis of skin lesions remains a ch.llenging *** lesion is a common type of skin cancer th.t exists *** is one of th. latest tech.ologies used for th. diagnosis of skin ***:Many computerized meth.ds h.ve been introduced in th. literature to classify skin ***,ch.llenges remain such.as imbalanced datasets,low contrast lesions,and th. extraction of irrelevant or redundant *** Work:In th.s study,a new tech.ique is proposed based on th. conventional and deep learning *** proposed framework consists of two major tasks:lesion segmentation and *** th. lesion segmentation task,contrast is initially improved by th. fusion of two filtering tech.iques and th.n performed a color transformation to color lesion area color ***,th. best ch.nnel is selected and th. lesion map is computed,wh.ch.is furth.r converted into a binary form using a th.esh.lding *** th. lesion classification task,two pre-trained CNN models were modified and trained using transfer *** features were extracted from both.models and fused using canonical correlation *** th. fusion process,a few redundant features were also added,lowering classification accuracy.A new tech.ique called maximum entropy score-based selection(MESbS)is proposed as a solution to th.s *** features selected th.ough.th.s approach.are fed into a cubic support vector mach.ne(C-SVM)for th. final ***:Th. experimental process was conducted on two datasets:ISIC 2017 and *** ISIC 2017 dataset was used for th. lesion segmentation task,wh.reas th. h.M10000 dataset was used for th. classification *** ach.eved accuracy for both.datasets was 95.6% and 96.7%, respectively, wh.ch.was h.gh.r th.nth. existing tech.iques.

关键词： Skin cancer lesion segmentation deep learning features fusion classification

Novel Optimized Cross-Layer Design with.Maximum Weigh.ed Capacity Based Resource Allocation for AMC/h.RQ Wireless Networks

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Wireless Engineering and Tech.ology 2013年第2期4卷 77-86页

作者： reham A. El-Mayet h.sh.m M. El-Badawy Salwa h. Elramly Electronic and Communication Department Ain Shams University Cairo Egypt National Telecommunication Institute Cairo Egypt

To provide quality-of service (QoS) guarantees for h.terogeneous applications, most recent wireless communications tech.ologies and standards combine th. error-correcting capability of h.brid automatic repeat request (h.RQ) sch.mes at th. data link layer (DLL) with.th. adaptation ability of th. adaptive modulation and coding (AMC) modes at th. ph.sical layer (Ph.) layer. Th.s paper aims to investigate th. aggregated system capacity as well as th. breakdown of th.s capacity for different ACM modes in each.h.RQ sch.me. Th.s investigation was done by using maximum weigh.ed capacity (MWC) resource allocation at th. Ph. layer in conjunction with.a novel packet error rate (PER)-based sch.duling at th. medium access control (MAC) layer. As a result, th. dominant AMC mode corresponding to ch.nnel SNR was available.

关键词： AMC Cross-Layer Design h.RQ Nakagami-m Fading Ch.nnel QoS