Objective To evaluate the clinical application of atlantoaxial joint fusion using anterior transarticular screw fixation and bone grafting for atlantoaxial joint instability. Methods Twenty-three cases of atlantoaxial...Objective To evaluate the clinical application of atlantoaxial joint fusion using anterior transarticular screw fixation and bone grafting for atlantoaxial joint instability. Methods Twenty-three cases of atlantoaxial joint instability were展开更多
Multimodal Sentiment Analysis(SA)is gaining popularity due to its broad application potential.The existing studies have focused on the SA of single modalities,such as texts or photos,posing challenges in effectively h...Multimodal Sentiment Analysis(SA)is gaining popularity due to its broad application potential.The existing studies have focused on the SA of single modalities,such as texts or photos,posing challenges in effectively handling social media data with multiple modalities.Moreover,most multimodal research has concentrated on merely combining the two modalities rather than exploring their complex correlations,leading to unsatisfactory sentiment classification results.Motivated by this,we propose a new visualtextual sentiment classification model named Multi-Model Fusion(MMF),which uses a mixed fusion framework for SA to effectively capture the essential information and the intrinsic relationship between the visual and textual content.The proposed model comprises three deep neural networks.Two different neural networks are proposed to extract the most emotionally relevant aspects of image and text data.Thus,more discriminative features are gathered for accurate sentiment classification.Then,a multichannel joint fusion modelwith a self-attention technique is proposed to exploit the intrinsic correlation between visual and textual characteristics and obtain emotionally rich information for joint sentiment classification.Finally,the results of the three classifiers are integrated using a decision fusion scheme to improve the robustness and generalizability of the proposed model.An interpretable visual-textual sentiment classification model is further developed using the Local Interpretable Model-agnostic Explanation model(LIME)to ensure the model’s explainability and resilience.The proposed MMF model has been tested on four real-world sentiment datasets,achieving(99.78%)accuracy on Binary_Getty(BG),(99.12%)on Binary_iStock(BIS),(95.70%)on Twitter,and(79.06%)on the Multi-View Sentiment Analysis(MVSA)dataset.These results demonstrate the superior performance of our MMF model compared to single-model approaches and current state-of-the-art techniques based on model evaluation criteria.展开更多
The fractal dimension of the fusion line in different dissimilar welded joints is measured with Box Dimension Method.The non scale region of the fusion line with fractal character is calculated. The fusion line in th...The fractal dimension of the fusion line in different dissimilar welded joints is measured with Box Dimension Method.The non scale region of the fusion line with fractal character is calculated. The fusion line in the dissimilar welded joint is proved to be a fractal structure. The change and influence factors of the fractal dimension of the fusion line are studied.展开更多
A new view of characterstic zone classification of fusion welding joint has been put for-ward on the base of a number of metallograplic observations and researches. TLe characteristiczones of the joint include (1) hom...A new view of characterstic zone classification of fusion welding joint has been put for-ward on the base of a number of metallograplic observations and researches. TLe characteristiczones of the joint include (1) homogenous mixture region (2)heterogerous mixture zone, (3)partically melting zone and (4) heat-affected zone. (1) and (2) consist of the weld metal. (2) and (3)compose the bond, the boundary betweer (2) and (3) is the fusion line Four kinds of characteristicappearences in the ' heterogenous mixture zone' are induced. The formation process of thecharcteristic zones is distussed in detail. The differences between authors' classification and W. F.Savage's one are compared, to hoping that the formation essence and composition feature of fusionwelding joint can be reasonably reflected.展开更多
中文电子病历实体关系抽取是构建医疗知识图谱,服务下游子任务的重要基础。目前,中文电子病例进行实体关系抽取仍存在因医疗文本关系复杂、实体密度大而造成医疗名词识别不准确的问题。针对这一问题,提出了基于对抗学习与多特征融合的...中文电子病历实体关系抽取是构建医疗知识图谱,服务下游子任务的重要基础。目前,中文电子病例进行实体关系抽取仍存在因医疗文本关系复杂、实体密度大而造成医疗名词识别不准确的问题。针对这一问题,提出了基于对抗学习与多特征融合的中文电子病历实体关系联合抽取模型AMFRel(adversarial learning and multi-feature fusion for relation triple extraction),提取电子病历的文本和词性特征,得到融合词性信息的编码向量;利用编码向量联合对抗训练产生的扰动生成对抗样本,抽取句子主语;利用信息融合模块丰富文本结构特征,并根据特定的关系信息抽取出相应的宾语,得到医疗文本的三元组。采用CHIP2020关系抽取数据集和糖尿病数据集进行实验验证,结果显示:AMFRel在CHIP2020关系抽取数据集上的Precision为63.922%,Recall为57.279%,F1值为60.418%;在糖尿病数据集上的Precision、Recall和F1值分别为83.914%,67.021%和74.522%,证明了该模型的三元组抽取性能优于其他基线模型。展开更多
目的:比较类风湿性关节炎(rheumatoid arthritis,RA)前足畸形重建术第5跖骨头切除成形与不切除的临床疗效。方法:回顾性分析2015年5月至2019年1月收治的50例(76足)女性RA致前足中重度畸形患者,依据第5跖趾关节退变情况,行RA前足重建第5...目的:比较类风湿性关节炎(rheumatoid arthritis,RA)前足畸形重建术第5跖骨头切除成形与不切除的临床疗效。方法:回顾性分析2015年5月至2019年1月收治的50例(76足)女性RA致前足中重度畸形患者,依据第5跖趾关节退变情况,行RA前足重建第5跖骨头保留或切除术,分为第5跖骨头保留组(保留组)和第5跖骨头切除组(切除组)。保留组24例,年龄47~81(60.37±8.60)岁;病程13~22(19.00±3.06)年;身体质量指数(body mass index,BMI)21~28(23.53±2.47)kg·m-2;中度拇外翻畸形6例(6足),重度拇外翻畸形18例(30足);采用第1跖趾关节融合联合第2-4跖趾关节成形及第5跖趾关节清理术治疗。切除组26例,年龄30~80(58.53±13.70)岁;病程8~25(17.94±3.92)年;BMI为20~28(24.60±2.03)kg·m-2;中度拇外翻畸形4例(4足),重度拇外翻畸形22例(36足);采用第1跖趾关节融合联合2-5跖趾关节跖骨头切除成形术治疗。观察并比较两组手术时间、术后并发症情况,比较术前和末次随访时拇外翻角(hallux valgus angle,HVA),第1-2跖骨间角(intermetatarsal angles between the first and the second metatarsals,IMAFS),第1-5跖骨间角(intermetatarsal angles between the first and the fifth metatarsals,IMAFF)变化情况,并采用日本足踝外科协会(Japanese Society for Surgery of Foot,JSSF)评价临床疗效。结果:50例均获得随访,切除组随访时间为14~46(25.30±8.83)个月;保留组随访时间为12~48(24.30±11.12)个月,两组随访时间比较,差异无统计学意义(P>0.05)。两组手术时间、术后并发症比较,差异无统计学意义(P>0.05)。切除组JSSF评分、HVA、IMAFS、IMAFF分别由术前的(45.09±3.35)分、(38.90±13.67)°、(12.88±1.72)°、(32.50±2.99)°改善至末次随访时的(81.60±3.27)分、(15.40±0.96)°、(9.06±2.27)°、(22.20±1.98)°(P<0.05);保留组分别由(47.09±3.96)分、(43.30±12.65)°、(13.99±3.13)°、(33.20±6.14)°改善至(83.10±3.66)分、(15.20±1.54)°、(8.99±1.02)°、(24.70±1.88)°。两组间术前及末次随访时JSSF评分、HVA、IMAFS、IMAFF比较,差异均无统计学意义(P>0.05)。末次随访时,两组JSSF评分中疼痛、畸形比较,差异有统计学意义(P<0.05)。结论:无论是类风湿前足重建术还是前足重建第5跖趾关节清理术,术后临床疗效及影像学结果均明显改善。对比类风湿前足重建术,类风湿前足重建第5跖趾关节清理术治疗类风湿前足中重度畸形,疼痛改善明显,但畸形改善较差。对于类风湿引起的前足中重度畸形,第5跖趾头关节面轻中度退变畸形的患者可考虑使用。展开更多
为获得结构化的小麦品种表型和遗传描述,针对非结构化小麦种质数据中存在的实体边界模糊以及关系重叠问题,提出一种基于深度字词融合的小麦种质信息实体关系联合抽取模型WGIE-DCWF(wheat germplasm information extraction model based ...为获得结构化的小麦品种表型和遗传描述,针对非结构化小麦种质数据中存在的实体边界模糊以及关系重叠问题,提出一种基于深度字词融合的小麦种质信息实体关系联合抽取模型WGIE-DCWF(wheat germplasm information extraction model based on deep character and word fusion)。模型编码层通过深度字词融合和上下文语义特征融合,提高密集实体特征识别能力;模型三元组抽取层建立层叠指针网络,提高重叠关系的提取能力。在小麦种质数据集和公开数据集上的一系列对比实验结果表明,WGIE-DCWF模型能够有效提高小麦种质数据实体关系联合抽取效果,同时拥有较好的泛化性,可以为小麦种质信息知识库构建提供技术支撑。展开更多
目的:探讨后路寰枢椎侧块关节cage植骨融合内固定术治疗难复性寰枢椎脱位的临床疗效,并与经口咽松解后路复位固定融合术进行疗效对比。方法:回顾性分析2018年1月~2022年8月我科采用后路寰枢椎侧块关节cage植骨融合内固定术(23例,cage组...目的:探讨后路寰枢椎侧块关节cage植骨融合内固定术治疗难复性寰枢椎脱位的临床疗效,并与经口咽松解后路复位固定融合术进行疗效对比。方法:回顾性分析2018年1月~2022年8月我科采用后路寰枢椎侧块关节cage植骨融合内固定术(23例,cage组)与经口咽松解后路复位固定融合术(25例,对照组)治疗的难复性寰枢椎脱位患者的临床资料,cage组男8例,女15例,年龄9~79岁(48.35±14.38岁);对照组男6例,女19例,年龄21~69岁(47.84±13.51岁)。记录两组患者手术时间、术中出血量、住院时间及并发症情况,术前、术后及末次随访时使用JOA评分评估患者神经功能状态,测量术前、术后及末次随访时的寰齿间距(atlantodental interval,ADI)、齿状突顶点距离Chamberlain线的垂直距离(vertical distance from odon to idprocess to Chamberlain′s line,DOCL)、延髓颈髓角(cervicomedullary angle,CMA)、斜坡枢椎角(clivus-axial angle,CAA),评估寰枢椎复位情况。评估侧块关节cage及后方植骨融合情况。结果:所有患者内固定位置良好,减压充分复位满意,症状均明显缓解,未出现椎动脉损伤和脊髓损伤加重。cage组手术时间133.04±34.04min、术中出血量119.13±54.77mL、住院时间14.74±6.10d,均明显短于或少于对照组(253.20±53.98min、181.20±45.40mL、23.96±5.47d)。cage组术前JOA、ADI、DOCL、CMA、CAA分别为6.33±1.13分、7.31±3.05mm、9.47±3.32mm、122.89°±12.58°、122.02°±12.50°,术后分别为13.04±2.17分、2.18±0.67mm、0.89±1.00mm、148.81°±5.43°、146.70°±9.32°,末次随访时分别为14.89±1.17分、2.09±0.69mm、0.83±0.86mm、149.10°±5.11°、146.89°±8.95°;对照组术前JOA、ADI、DOCL、CMA、CAA分别为6.76±1.21分、7.70±0.97mm、10.56±1.99mm、121.53°±4.87°、123.77°±8.95°,术后分别为13.26±1.32分、1.89±0.50mm、1.13±1.08mm、151.40°±6.15°、149.86°±5.58°,末次随访时分别为15.02±0.88分、1.87±0.44mm、0.87±1.39mm、149.48°±4.06°、149.94°±6.61°,两组术后及末次随访JOA、ADI、DOCL、CMA及CAA均较术前明显改善(P<0.05),术后JOA评分与末次随访相比存在统计学差异(P<0.05),但ADI、DOCL、CMA及CAA无统计学差异(P>0.05)。cage组仅1例切口感染;对照组3例切口感染(口咽2例,后路1例),1例脑脊液漏。两组随访期间内固定在位稳定,末次随访植骨均达到骨性融合,cage组关节间隙高度无丢失。结论:难复性寰枢椎脱位采用后路寰枢椎侧块关节cage植骨融合内固定术与经口咽松解后路复位固定融合术相比疗效相当,但增加了植骨融合位点,能更有效融合,避免了经口手术,减少了手术时间、术中出血量、住院时间及并发症的发生。展开更多
文摘Objective To evaluate the clinical application of atlantoaxial joint fusion using anterior transarticular screw fixation and bone grafting for atlantoaxial joint instability. Methods Twenty-three cases of atlantoaxial joint instability were
文摘Multimodal Sentiment Analysis(SA)is gaining popularity due to its broad application potential.The existing studies have focused on the SA of single modalities,such as texts or photos,posing challenges in effectively handling social media data with multiple modalities.Moreover,most multimodal research has concentrated on merely combining the two modalities rather than exploring their complex correlations,leading to unsatisfactory sentiment classification results.Motivated by this,we propose a new visualtextual sentiment classification model named Multi-Model Fusion(MMF),which uses a mixed fusion framework for SA to effectively capture the essential information and the intrinsic relationship between the visual and textual content.The proposed model comprises three deep neural networks.Two different neural networks are proposed to extract the most emotionally relevant aspects of image and text data.Thus,more discriminative features are gathered for accurate sentiment classification.Then,a multichannel joint fusion modelwith a self-attention technique is proposed to exploit the intrinsic correlation between visual and textual characteristics and obtain emotionally rich information for joint sentiment classification.Finally,the results of the three classifiers are integrated using a decision fusion scheme to improve the robustness and generalizability of the proposed model.An interpretable visual-textual sentiment classification model is further developed using the Local Interpretable Model-agnostic Explanation model(LIME)to ensure the model’s explainability and resilience.The proposed MMF model has been tested on four real-world sentiment datasets,achieving(99.78%)accuracy on Binary_Getty(BG),(99.12%)on Binary_iStock(BIS),(95.70%)on Twitter,and(79.06%)on the Multi-View Sentiment Analysis(MVSA)dataset.These results demonstrate the superior performance of our MMF model compared to single-model approaches and current state-of-the-art techniques based on model evaluation criteria.
文摘The fractal dimension of the fusion line in different dissimilar welded joints is measured with Box Dimension Method.The non scale region of the fusion line with fractal character is calculated. The fusion line in the dissimilar welded joint is proved to be a fractal structure. The change and influence factors of the fractal dimension of the fusion line are studied.
文摘A new view of characterstic zone classification of fusion welding joint has been put for-ward on the base of a number of metallograplic observations and researches. TLe characteristiczones of the joint include (1) homogenous mixture region (2)heterogerous mixture zone, (3)partically melting zone and (4) heat-affected zone. (1) and (2) consist of the weld metal. (2) and (3)compose the bond, the boundary betweer (2) and (3) is the fusion line Four kinds of characteristicappearences in the ' heterogenous mixture zone' are induced. The formation process of thecharcteristic zones is distussed in detail. The differences between authors' classification and W. F.Savage's one are compared, to hoping that the formation essence and composition feature of fusionwelding joint can be reasonably reflected.
文摘中文电子病历实体关系抽取是构建医疗知识图谱,服务下游子任务的重要基础。目前,中文电子病例进行实体关系抽取仍存在因医疗文本关系复杂、实体密度大而造成医疗名词识别不准确的问题。针对这一问题,提出了基于对抗学习与多特征融合的中文电子病历实体关系联合抽取模型AMFRel(adversarial learning and multi-feature fusion for relation triple extraction),提取电子病历的文本和词性特征,得到融合词性信息的编码向量;利用编码向量联合对抗训练产生的扰动生成对抗样本,抽取句子主语;利用信息融合模块丰富文本结构特征,并根据特定的关系信息抽取出相应的宾语,得到医疗文本的三元组。采用CHIP2020关系抽取数据集和糖尿病数据集进行实验验证,结果显示:AMFRel在CHIP2020关系抽取数据集上的Precision为63.922%,Recall为57.279%,F1值为60.418%;在糖尿病数据集上的Precision、Recall和F1值分别为83.914%,67.021%和74.522%,证明了该模型的三元组抽取性能优于其他基线模型。
文摘目的:比较类风湿性关节炎(rheumatoid arthritis,RA)前足畸形重建术第5跖骨头切除成形与不切除的临床疗效。方法:回顾性分析2015年5月至2019年1月收治的50例(76足)女性RA致前足中重度畸形患者,依据第5跖趾关节退变情况,行RA前足重建第5跖骨头保留或切除术,分为第5跖骨头保留组(保留组)和第5跖骨头切除组(切除组)。保留组24例,年龄47~81(60.37±8.60)岁;病程13~22(19.00±3.06)年;身体质量指数(body mass index,BMI)21~28(23.53±2.47)kg·m-2;中度拇外翻畸形6例(6足),重度拇外翻畸形18例(30足);采用第1跖趾关节融合联合第2-4跖趾关节成形及第5跖趾关节清理术治疗。切除组26例,年龄30~80(58.53±13.70)岁;病程8~25(17.94±3.92)年;BMI为20~28(24.60±2.03)kg·m-2;中度拇外翻畸形4例(4足),重度拇外翻畸形22例(36足);采用第1跖趾关节融合联合2-5跖趾关节跖骨头切除成形术治疗。观察并比较两组手术时间、术后并发症情况,比较术前和末次随访时拇外翻角(hallux valgus angle,HVA),第1-2跖骨间角(intermetatarsal angles between the first and the second metatarsals,IMAFS),第1-5跖骨间角(intermetatarsal angles between the first and the fifth metatarsals,IMAFF)变化情况,并采用日本足踝外科协会(Japanese Society for Surgery of Foot,JSSF)评价临床疗效。结果:50例均获得随访,切除组随访时间为14~46(25.30±8.83)个月;保留组随访时间为12~48(24.30±11.12)个月,两组随访时间比较,差异无统计学意义(P>0.05)。两组手术时间、术后并发症比较,差异无统计学意义(P>0.05)。切除组JSSF评分、HVA、IMAFS、IMAFF分别由术前的(45.09±3.35)分、(38.90±13.67)°、(12.88±1.72)°、(32.50±2.99)°改善至末次随访时的(81.60±3.27)分、(15.40±0.96)°、(9.06±2.27)°、(22.20±1.98)°(P<0.05);保留组分别由(47.09±3.96)分、(43.30±12.65)°、(13.99±3.13)°、(33.20±6.14)°改善至(83.10±3.66)分、(15.20±1.54)°、(8.99±1.02)°、(24.70±1.88)°。两组间术前及末次随访时JSSF评分、HVA、IMAFS、IMAFF比较,差异均无统计学意义(P>0.05)。末次随访时,两组JSSF评分中疼痛、畸形比较,差异有统计学意义(P<0.05)。结论:无论是类风湿前足重建术还是前足重建第5跖趾关节清理术,术后临床疗效及影像学结果均明显改善。对比类风湿前足重建术,类风湿前足重建第5跖趾关节清理术治疗类风湿前足中重度畸形,疼痛改善明显,但畸形改善较差。对于类风湿引起的前足中重度畸形,第5跖趾头关节面轻中度退变畸形的患者可考虑使用。
文摘为获得结构化的小麦品种表型和遗传描述,针对非结构化小麦种质数据中存在的实体边界模糊以及关系重叠问题,提出一种基于深度字词融合的小麦种质信息实体关系联合抽取模型WGIE-DCWF(wheat germplasm information extraction model based on deep character and word fusion)。模型编码层通过深度字词融合和上下文语义特征融合,提高密集实体特征识别能力;模型三元组抽取层建立层叠指针网络,提高重叠关系的提取能力。在小麦种质数据集和公开数据集上的一系列对比实验结果表明,WGIE-DCWF模型能够有效提高小麦种质数据实体关系联合抽取效果,同时拥有较好的泛化性,可以为小麦种质信息知识库构建提供技术支撑。
文摘目的:探讨后路寰枢椎侧块关节cage植骨融合内固定术治疗难复性寰枢椎脱位的临床疗效,并与经口咽松解后路复位固定融合术进行疗效对比。方法:回顾性分析2018年1月~2022年8月我科采用后路寰枢椎侧块关节cage植骨融合内固定术(23例,cage组)与经口咽松解后路复位固定融合术(25例,对照组)治疗的难复性寰枢椎脱位患者的临床资料,cage组男8例,女15例,年龄9~79岁(48.35±14.38岁);对照组男6例,女19例,年龄21~69岁(47.84±13.51岁)。记录两组患者手术时间、术中出血量、住院时间及并发症情况,术前、术后及末次随访时使用JOA评分评估患者神经功能状态,测量术前、术后及末次随访时的寰齿间距(atlantodental interval,ADI)、齿状突顶点距离Chamberlain线的垂直距离(vertical distance from odon to idprocess to Chamberlain′s line,DOCL)、延髓颈髓角(cervicomedullary angle,CMA)、斜坡枢椎角(clivus-axial angle,CAA),评估寰枢椎复位情况。评估侧块关节cage及后方植骨融合情况。结果:所有患者内固定位置良好,减压充分复位满意,症状均明显缓解,未出现椎动脉损伤和脊髓损伤加重。cage组手术时间133.04±34.04min、术中出血量119.13±54.77mL、住院时间14.74±6.10d,均明显短于或少于对照组(253.20±53.98min、181.20±45.40mL、23.96±5.47d)。cage组术前JOA、ADI、DOCL、CMA、CAA分别为6.33±1.13分、7.31±3.05mm、9.47±3.32mm、122.89°±12.58°、122.02°±12.50°,术后分别为13.04±2.17分、2.18±0.67mm、0.89±1.00mm、148.81°±5.43°、146.70°±9.32°,末次随访时分别为14.89±1.17分、2.09±0.69mm、0.83±0.86mm、149.10°±5.11°、146.89°±8.95°;对照组术前JOA、ADI、DOCL、CMA、CAA分别为6.76±1.21分、7.70±0.97mm、10.56±1.99mm、121.53°±4.87°、123.77°±8.95°,术后分别为13.26±1.32分、1.89±0.50mm、1.13±1.08mm、151.40°±6.15°、149.86°±5.58°,末次随访时分别为15.02±0.88分、1.87±0.44mm、0.87±1.39mm、149.48°±4.06°、149.94°±6.61°,两组术后及末次随访JOA、ADI、DOCL、CMA及CAA均较术前明显改善(P<0.05),术后JOA评分与末次随访相比存在统计学差异(P<0.05),但ADI、DOCL、CMA及CAA无统计学差异(P>0.05)。cage组仅1例切口感染;对照组3例切口感染(口咽2例,后路1例),1例脑脊液漏。两组随访期间内固定在位稳定,末次随访植骨均达到骨性融合,cage组关节间隙高度无丢失。结论:难复性寰枢椎脱位采用后路寰枢椎侧块关节cage植骨融合内固定术与经口咽松解后路复位固定融合术相比疗效相当,但增加了植骨融合位点,能更有效融合,避免了经口手术,减少了手术时间、术中出血量、住院时间及并发症的发生。
