目的 针对计算机断层扫描(Computed Tomography,CT)中由于不完全扫描数据导致图像噪声伪影严重的问题,提出一种基于Transformer块的混合域网络稀疏角度CT成像算法(Hybrid Domain network for sparse view CT imaging based on Transform...目的 针对计算机断层扫描(Computed Tomography,CT)中由于不完全扫描数据导致图像噪声伪影严重的问题,提出一种基于Transformer块的混合域网络稀疏角度CT成像算法(Hybrid Domain network for sparse view CT imaging based on Transformer,HDTransformer)。方法 算法的主要思想是借助于新型的Transformer网络,构建适用于多阶段稀疏角度CT投影数据及图像数据的处理流,以提高稀疏角度CT图像重建质量;与现有两阶段混合域处理方法相比,本方法采用图像域-投影域-图像域三阶段混合处理流程,通过多阶段信息的联合互补提高成像质量;此外,针对不同阶段数据噪声伪影特点设计不同的Transformer块,以实现差异化的处理;更进一步,算法采用可微分的解析重建和投影运算,建立投影域与图像域数据的转换,最终实现端到端的稀疏角度CT优质成像流。结果 通过Mayo数据实验验证,其视觉结果表明:处理后的不同部位CT图像噪声伪影均能够得到较好的抑制;量化结果表明:处理后的CT图像峰值信噪比和特征相似性均优于对比方法。结论 实验的定性和定量结果表明:所提算法在去除图像伪影噪声方面要优于其他算法,具有更高的质量,验证了该方法的有效性。展开更多
Automated writing evaluation is the use of computer programs to evaluate essays written in an educational setting.This study aims to evaluate the state of the art and the trends of automated writing evaluation researc...Automated writing evaluation is the use of computer programs to evaluate essays written in an educational setting.This study aims to evaluate the state of the art and the trends of automated writing evaluation research.A total of 206 automated writing evaluation-related publications published in 88 journals between 1995 and 2022 were used as the data in our study.Bibliometric methods were employed to identify the publication trends,venues for publication,productive authors,highly cited articles,seminal works,and the most frequently explored topics across the past decades.The findings show that automated writing evaluation has become a fast-growing interdisciplinary area of interest.In addition,this line of research has witnessed a growing interest in topics pertaining to the latest technology,the perception of children,the teachers’feedback,and complex tasks.Implications of our study for future research are also discussed.展开更多
文摘Automated writing evaluation is the use of computer programs to evaluate essays written in an educational setting.This study aims to evaluate the state of the art and the trends of automated writing evaluation research.A total of 206 automated writing evaluation-related publications published in 88 journals between 1995 and 2022 were used as the data in our study.Bibliometric methods were employed to identify the publication trends,venues for publication,productive authors,highly cited articles,seminal works,and the most frequently explored topics across the past decades.The findings show that automated writing evaluation has become a fast-growing interdisciplinary area of interest.In addition,this line of research has witnessed a growing interest in topics pertaining to the latest technology,the perception of children,the teachers’feedback,and complex tasks.Implications of our study for future research are also discussed.
