|
|
|
题名
|
作者
|
年代
|
出处
|
被引量
|
| 1 | Development and validation of an endoscopic images-based deep learning model for detection with nasopharyngeal malignancies显示文摘Background:Due to the occult anatomic location of the nasopharynx and frequent presence of adenoid hyperpla-sia,the positive rate for malignancy identification during biopsy is low,thus leading to delayed or missed diagnosis for nasopharyngeal malignancies upon initial attempt.Here,we aimed to develop an artificial intelligence tool to detect nasopharyngeal malignancies under endoscopic examination based on deep learning.Methods:An endoscopic images-based nasopharyngeal malignancy detection model(eNPM-DM)consisting of a fully convolutional network based on the inception architecture was developed and fine-tuned using separate training and validation sets for both classification and segmentation.Briefly,a total of 28,966 qualified images were collected.Among these images,27,536 biopsy-proven images from 7951 individuals obtained from January 1st,2008,to December 31st,2016,were split into the training,validation and test sets at a ratio of 7:1:2 using simple randomiza-tion.Additionally,1430 images obtained from January 1st,2017,to March 31st,2017,were used as a prospective test set to compare the performance of the established model against oncologist evaluation.The dice similarity coef-ficient(DSC)was used to evaluate the efficiency of eNPM-DM in automatic segmentation of malignant area from the background of nasopharyngeal endoscopic images,by comparing automatic segmentation with manual segmenta-tion performed by the experts.Results:All images were histopathologically confirmed,and included 5713(19.7%)normal control,19,107(66.0%)nasopharyngeal carcinoma(NPC),335(1.2%)NPC and 3811(13.2%)benign diseases.The eNPM-DM attained an overall accuracy of 88.7%(95%confidence interval(CI)87.8%-89.5%)in detecting malignancies in the test set.In the prospective comparison phase,eNPM-DM outperformed the experts:the overall accuracy was 88.0%(95%CI 86.1%-89.6%)vs.80.5%(95%CI 77.0%-84.0%).The eNPM-DM required less time(40 s vs.110.0±5.8 min)and exhibited encouraging performance in automatic segmentation of nasopharyngeal malignant area from the background,with an average DSC of 0.78±0.24 and 0.75±0.26 in the test and prospective test sets,respectively.Conclusions:The eNPM-DM outperformed oncologist evaluation in diagnostic classification of nasopharyngeal mass into benign versus malignant,and realized automatic segmentation of malignant area from the background of nasopharyngeal endoscopic images. | Chaofeng Li Bingzhong Jing Liangru Ke Bin Li Weixiong Xia Caisheng He Chaonan Qian Chong Zhao Haiqiang Mai Mingyuan Chen Kajia Cao Haoyuan Mo Ling Guo Qiuyan Chen Linquan Tang Wenze Qiu Yahui Yu Hu Liang Xinjun Huang Guoying Liu Wangzhong Li Lin Wang Rui Sun Xiong Zou Shanshan Guo Peiyu Huang Donghua Luo Fang Qiu Yishan Wu Yijun Hua Kuiyuan Liu Shuhui Lv Jingjing Miao Yanqun Xiang Ying Sun Xiang Guo Xing Lv | 2018 | Cancer Communications2018，38，1： | 9 |
| 2 | Bacillus cereus AR156 primes induced systemic resistance by suppressing miR825/825 and activating defense-related genes in Arabidopsis显示文摘Small RNAs play an important role in plant immune responses. However, their regulatory function in induced systemic resistance(ISR) is nascent. Bacillus cereus AR156 is a plant growth-promoting rhizobacterium that induces ISR in Arabidopsis against bacterial infection. Here,by comparing small RNA profiles of Pseudomonas syringae pv. tomato(Pst) DC3000-infected Arabidopsis with and without AR156 pretreatment, we identified a group of Arabidopsis micro RNAs(mi RNAs) that are differentially regulated by AR156 pretreatment. mi R825 and mi R825 are two mi RNA generated from a single mi RNA gene.Northern blot analysis indicated that they were significantly downregulated in Pst DC3000-infected plants pretreated with AR156, in contrast to the plants without AR156 pretreatment. mi R825 targets two ubiquitin-protein ligases,while mi R825 targets toll-interleukin-like receptor(TIR)-nucleotide binding site(NBS) and leucine-rich repeat(LRR)type resistance(R) genes. The expression of these target genes negatively correlated with the expression of mi R825 and mi R825. Moreover, transgenic plants showing reduced expression of mi R825 and mi R825 displayed enhanced resistance to Pst DC3000 infection, whereas transgenic plants overexpressing mi R825 and mi R825 were more susceptible. Taken together, our data indicates that Bacillus cereus AR156 pretreatment primes ISR to Pst infection by suppressing mi R825 and mi R825 and activating the defense related genes they targeted. | Dongdong Niu Jing Xia Chunhao Jiang Beibei Qi Xiaoyu Ling Siyuan Lin Weixiong Zhang Jianhua Guo Hailing Jin Hongwei Zhao | 2016 | Journal of Integrative Plant Biology2016，58，4： | 6 |
| 3 | Reducing active layer thickness of polyamide composite membranes using a covalent organic framework interlayer in interfacial polymerization显示文摘Polyamide(PA)-based thin-film composite membranes exhibit enormous potential in water purification,owing to their facile fabrication,decent performance and desirable stability.However,the thick PA active layer with high transport resistance from the conventional interfacial polymerization hampers their applications.The controllable fabrication of a thin PA active layer is essential for high separation efficiency but still challenging.Herein,a covalent organic framework TpPa-1 interlayer was firstly deposited on a polyethersulfone(PES)substrate to reduce the thickness of PA active layer in interfacial polymerization.The abundant pores of TpPa-1 increase the local concentration of amine monomers by adsorbing piperazine molecules,while hydrogen bonds between hydrophilic groups of TpPa-1 and piperazine molecules slow down their diffusion rate.Arising from those synergetic effects,the PA active layer is effectively reduced from 200 nm to 120 nm.By optimizing TpPa-1 interlayer and PA active layer,the water flux of resultant membranes can reach 171.35 L·m^-2·h^-1·MPa^-1,which increased by 125.4%compared with PA/PES membranes,while the rejection rates of sodium sulfate and dyes solution remained more than 90%and 99%,respectively.Our strategy may stimulate rational design of ultrathin PA-based nanofiltration membranes with high performances. | Meidi Wang Weixiong Guo Zhongyi Jiang Fusheng Pan | 2020 | Chinese Journal of Chemical Engineering2020，28，4： | 5 |
| 4 | Characterization of secondary organic aerosol particles using aerosol laser time-of-flight mass spectrometer coupled with FCM clustering algorithm显示文摘 | Mingqiang Huang Liqing Hao Xiaoyong Guo Changjin Hu Xuejun Gu Weixiong Zhao Zhenya Wang Li Fang Weijun Zhang | 2013 | Atmospheric Environment2013，，： | 1 |
| 5 | Effect of CeO_(2) on carbon deposition resistance of Ni/CeO_(2) catalyst supported on SiC porous ceramic for ethanol steam reforming显示文摘Ni/CeO_(2) catalysts(nCeO_(2):n_(Ni)=0,1,4,7,10)supported on SiC porous ce ramics for ethanol steam reforming(ESR)were investigated with respect to hydrogen production performance and growth of carbon deposition.The oxygen released from CeO_(2) enables the oxidation of CH_(x) species to serve as carbon precursors,thus providing Ni/CeO_(2) catalysts with stronger resistance to carbon deposition compared with Ni catalysts.The Ni/CeO_(2) catalysts prepared by inverse microemulsion and impregnation methods exhibit regular semicircular spherical shape on SiC porous ceramics.Under 500℃for 25 h of ESR reaction,the ethanol conversion rate over Ni/CeO_(2) catalysts(n_(CeO_(2)):n_(Ni)=7)is sustained up to 100%and H_(2) selectivity is essentially kept at 74%.The by-product selectivity declines stepwise with increasing content of CeO_(2),which is attributed to the adsorption and oxidation of CO and of CH_(x) species as CH_4 precursor from CeO_(2).The scanning electron microscopy(SEM)and transform electron microscopy(TEM)results reveal that further loading of CeO_(2) on the surface of Ni catalysts can alleviate both migration and sintering of Ni particles.Furthermore,carbon deposition on Ni/CeO_(2) catalysts preferentially outgrow filamentous rather than amorphous carbon,with a tendency for the latter to be more deactivated. | Tingting Bao Hui zhou Yi Zhang Chenxu Guo Wenming Guo Hang Qin Pengzhao Gao Hanning Xiao Weixiong Liu | 2023 | Journal of Rare Earths2023，41，11： | 0 |
| 6 | 基于内镜图像深度学习的鼻咽恶性肿瘤检测模型的建立与验证显示文摘背景与目的由于鼻咽部解剖位置隐匿且腺体增生频发,活检时恶性肿瘤的阳性率较低,从而导致初诊时鼻咽恶性肿瘤确诊延时或漏诊。本文旨在建立一种人工智能工具——基于深度学习的内镜检查,来检测鼻咽恶性肿瘤。方法建立了一种基于内镜图像的鼻咽恶性肿瘤检测模型(endoscopic imagesbased nasopharyngeal malignancies detection model,eNPM-DM),该模型由基于空间结构的全卷积网络构成,采用单独训练集和验证集对分类和分割进行微调。总共收集了28,966张合格图像。其中,自2008年1月1日至2016年12月31日,从7951例个体中获得了27,536张经活检证实的图像,按照7∶1∶2的比例随机分为训练、验证和测试集。此外,将2017年1月1日到2017年3月31日获得的1430张图像纳入预测集,用以对建立模型的性能与肿瘤专家的评价进行比较。以鼻咽镜图像为背景,对自动分割和专家手工分割进行比较,采用dice相似系数(dice similarity coefficient,DSC)评价eNPM-DM从鼻咽部内镜图像的背景中自动分割出恶性肿瘤区域的效率。结果所有图像经过病理组织学验证,包括正常对照5713(19.7%)例、鼻咽癌(nasopharyngeal carcinoma,NPC)19,107(66.0%)例、其他恶性肿瘤335(1.2%)例和3811(13.2%)例良性病变。在测试集中,eNPM-DM检测恶性肿瘤的总准确率达88.7%[95%置信区间(confidence interval,CI):87.8%–89.5%]。在预测比较阶段,eNPM-DM表现优于专家:总准确率分别为88.0%(95%CI:86.1%–89.6%)和80.5%(95%CI:77.0%–84.0%)。eNPM-DM耗时更短(40 s vs. 110.0±5.8 min),且从背景中自动分割出鼻咽恶性肿瘤区域方面表现优秀,测试集和预测集中的平均DSC分别为0.78±0.24和0.75±0.26。结论 eNPM-DM在鼻咽肿块良性/恶性诊断分类方面优于肿瘤学家评估,并且实现了从鼻咽内镜图像背景中对恶性区域自动分割。 | Chaofeng Li Bingzhong Jing Liangru Ke Bin Li Weixiong Xia Caisheng He Chaonan Qian Chong Zhao Haiqiang Mai Mingyuan Chen Kajia Cao Haoyuan Mo Ling Guo Qiuyan Chen Linquan Tang Wenze Qiu Yahui Yu Hu Liang Xinjun Huang Guoying Liu Wangzhong Li Lin Wang Rui Sun Xiong Zou Shanshan Guo Peiyu Huang Donghua Luo Fang Qiu Yishan Wu Yijun Hua Kuiyuan Liu Shuhui Lv Jingjing Miao Yanqun Xiang Ying Sun Xiang Guo Xing Lv | 2019 | 癌症2019，38，7： | 0 |
您所在的机构暂未开通此部分的功能，请联系管理员