Editors Recommend
20 September 2020, Volume 6 Issue 5
TOPIC:MEDICAL BIG DATA
An automatic ICD coding method for clinical records based on deep neural network
Yichao DU,Tong XU,Jianhui MA,Enhong CHEN,Yi ZHENG,Tongzhu LIU,Guixian TONG
2020, 6(5):  0.  doi:10.11959/j.issn.2096-0271.2020040
Asbtract ( 38 )   HTML ( 15)   PDF (1379KB) ( 12 )   Knowledge map   
Figures and Tables | References | Related Articles | Metrics

With the increase in the number of the international classification of diseases (ICD) codes,the difficulty and cost of manual coding based on clinical records have greatly increased,and automatic ICD coding technology has attracted widespread attention.A multi-scale residual graph convolution network automatic ICD coding technology was proposed.This technology uses a multi-scale residual network to capture text patterns of different lengths of clinical text and extracts the hierarchical relationship between labels based on the graph convolutional neural network to enhance the ability of automatic coding.The experimental results on the real medical data set MIMIC-III show that the P@k and Micro-F1 of this method are 72.2% and 53.9%,respectively,which significantly improves the prediction performance.

Parallel optimization of variation detection algorithms for large-scale genome data
Yingbo CUI,Chun HUANG,Tao TANG,Canqun YANG,Xiangke LIAO,Shaoliang PENG
2020, 6(5):  0.  doi:10.11959/j.issn.2096-0271.2020041
Asbtract ( 26 )   HTML ( 16)   PDF (1331KB) ( 7 )   Knowledge map   
Figures and Tables | References | Related Articles | Metrics

Sequence alignment and mutation detection are the basic steps of genomic data analysis.They are the premise of subsequent functional analysis,and the most time-consuming steps.In order to effectively deal with the massive genomic big data brought by high-throughput sequencing technology,MPI,OpenMP and other technologies to perform multi-level parallel optimization of sequence alignment algorithm and SNP detection algorithm were used.By testing on different data sets and parallel scales,the core algorithm reached more than 9x speedup,and the parallel efficiency remained above 60% in large-scale test.The improved algorithms obtain good parallel performance and scalability,that effectively improves the ability of genomic big data mutation detection.

Application of medical big data in learning health system
Yangfan CHAI,Guilan KONG,Luxia ZHANG
2020, 6(5):  0.  doi:10.11959/j.issn.2096-0271.2020042
Asbtract ( 49 )   HTML ( 21)   PDF (1151KB) ( 44 )   Knowledge map   
Figures and Tables | References | Related Articles | Metrics

The learning health system (LHS) aims at accelerating the process of knowledge generation,transformation and application in clinical practice.Applying medical big data in LHS to meet the knowledge needs of patients and healthcare decision makers would help to promote the development of precision medicine.Firstly,the current status of medical big data and LHS were reviewed,then the characteristics and challenges of applying medical big data in LHS were analyzed by refering to some typical application cases.Finally,the challenges faced by LHS in China were addressed and the prospect of applying medical big data to LHS in the future was provided.

Study on domain adaptation of medical data based on generative adversarial network
Hufei YU,Jingxi WEN,Jiang XIN,Yan TANG
2020, 6(5):  0.  doi:10.11959/j.issn.2096-0271.2020043
Asbtract ( 15 )   HTML ( 7)   PDF (1886KB) ( 10 )   Knowledge map   
Figures and Tables | References | Related Articles | Metrics

In the study of medical imaging aided diagnosis,researchers often collect a lot of training data coming from different hospitals (named variety fields).But because of the certain field has insufficient training data,the deep learning model would get very poor performance on the test data of this field.To mitigate this problem,a method to study domain adaptation of the difference between male and female brain images based on the generative adversarial network was proposed.The data distribution of different domains was learned and the key features were extracted by using the generative adversarial network,and then the differences between male and female brain images in different domains were studied based on the extracted key features.Experiments show that the method can also achieve more than 80% recognition accuracy in the domain with only a small amount of data involved in training.

STUDY
A survey on multi-source heterogeneous data processing methods in manufacturing process
Shichao CHEN,Chunyu CUI,Hua ZHANG,Ge MA,Fenghua ZHU,Xiuqin SHANG,Gang XIONG
2020, 6(5):  0.  doi:10.11959/j.issn.2096-0271.2020044
Asbtract ( 28 )   HTML ( 11)   PDF (3347KB) ( 26 )   Knowledge map   
Figures and Tables | References | Related Articles | Metrics

The effective processing and deep mining analysis of data can provide manufacturers with more effective production scheduling,equipment management and other strategies to affect production yield and efficiency.The processing methods and technologies of multi-source heterogeneous data in the manufacturing process were systematically reviewed.Firstly,the multi-source heterogeneous data content and classification in the manufacturing process was clarified.Secondly,the data processing methods and techniques applied in various stages of data collection,data integration and data analysis in the multi-source heterogeneous data processing were described.And the advantages and disadvantages of various techniques and their applications were analyzed.Finally,the multi-source heterogeneous data processing methods and techniques in the manufacturing process were summarized.And the challenges and development trends were pointed out.

Aspect sentiment analysis based on a hierarchical attention network
Ting SONG,Zhanwei CHEN,Haifeng YANG
2020, 6(5):  0.  doi:10.11959/j.issn.2096-0271.2020045
Asbtract ( 20 )   HTML ( 8)   PDF (1676KB) ( 14 )   Knowledge map   
Figures and Tables | References | Related Articles | Metrics

Aspect sentiment analysis based on deep learning is one of the hot spots in natural language processing.Aiming at aspect sentiment,a deep hierarchical attention network model based on aspect sentiment analysis was proposed.The local features of the text and the temporal relationship of different sentences were retained in model through the convolutional neural network,and the emotional features within and between sentences were obtained by using the layered long shortterm memory network (LSTM).Among them,specific aspects of information were added to LSTM and a dynamic control chain was designed to improve the traditional LSTM.A comparative experiment is conducted on the two data sets in SemEval 2014 and the Twitter data set.Compared with the traditional model,the accuracy of sentiment classification of the proposed model increases by about 3%.

Research on secure data sharing system based on blockchain
Yansong LIU,Qi XIA,Zhu LI,Hu XIA,Xiaosong ZHANG,Jianbin GAO
2020, 6(5):  0.  doi:10.11959/j.issn.2096-0271.2020046
Asbtract ( 28 )   HTML ( 15)   PDF (2262KB) ( 28 )   Knowledge map   
Figures and Tables | References | Related Articles | Metrics

In view of the increasing number of people in the digital interaction process of direct or indirect disclosure of privacy issues,the establishment of a set of data security sharing system based on blockchain network,the access control algorithm based on ciphertext-policy attribute-based encryption and the homomorphic encryption algorithm provides reliable chain data sharing were mainly researched.Data sharing on chain architecture for data sharing on the chain was proposed.Finally,simulation experiments were carried out,and the results of experimental data were analyzed.The work effectively solves the problem of malicious parties using the transaction transparency of blockchain for data analysis and ensures the privacy security of user data in the sharing process.

APPLICATION
On-chain witness and off-chain transmission trustworthy data sharing platform
Zhao ZHANG,Jixin TIAN,Cheqing JIN
2020, 6(5):  0.  doi:10.11959/j.issn.2096-0271.2020047
Asbtract ( 26 )   HTML ( 8)   PDF (1648KB) ( 15 )   Knowledge map   
Figures and Tables | References | Related Articles | Metrics

Blockchain system can build a trusted infrastructure for sharing data between multiple untrusted parties.However,directly uploading original shared data to blockchain is not suitable for large-scale data sharing scenarios.A data sharing architecture where data sharing request and response records are deposited on-chain and original data is transmitted securely off-chain was proposed.The architecture can alleviate the problems of system overload and privacy protection to a certain extent.Finally,with the increase of participating nodes and the data sharing requests and responses to be handled per second,the limitations in distributed storage,consensus protocol,smart contract execution,and query from light clients,directions for further research were proposed,in order to specify the technical bottlenecks that need to be further broken for the existing blockchain system applied to the field of data sharing.

FORUM
Comparative analysis between bank industry data governance guidelines and DCMM
Hong DAI,Qun ZHANG,Haolin LU,Junzhi BIN
2020, 6(5):  0.  doi:10.11959/j.issn.2096-0271.2020048
Asbtract ( 16 )   HTML ( 7)   PDF (1537KB) ( 18 )   Knowledge map   
Figures and Tables | References | Related Articles | Metrics

Recently,data governance got widespread attention of many industries.Some standards and polices had been published from country and industries,these files defined data governance concept and framework and boost data governance industry development.These files were analysed and some differences between them were identified to help people understand the trends and focus of data governance industry.At same time,some suggestions about how to implement DCMM in the bank industry were provided to help banks to better meet regulatory requirement and improve the maturity level of data management ability.

TREND
2020, Vol. 6 No.4 No.3 No.2 No.1
2019, Vol. 5 No.6 No.5 No.4 No.3 No.2 No.1
2018, Vol. 4 No.6 No.5 No.4 No.3 No.2 No.1
2017, Vol. 3 No.6 No.5 No.4 No.3 No.2 No.1
2016, Vol. 2 No.6 No.5 No.4 No.3 No.2 No.1
2015, Vol. 1 No.4 No.2 No.3 No.1
Most Download
Most Read
Most Cited