Research And Application Of Clinical Data Analysis For Diabetes

With the development of information construction process of the major hospitals, various production systems in the hospitals such as HIS(Hospital Information System), EMR(electronic medical record system) has accumulated a large-scale clinical data. This clinical big data has a far-reaching significance for improving the quality of clinical care.Diabetes as a chronic disease, has a long treatment cycle, easy to lead to a variety of complications such as kidney disease, eye disease, disease, and other characteristics. A large number of clinical data of patients with diabetes, including laboratory testing, clinical diagnosis and medical information, which implied patient population characteristics, the change trend of the disease, drug efficacy and other key information. In order to find out the hidden knowledge, the analysis and research of the clinical date is more and more important.For this, this paper designs and implements the clinical data analysis and application system of diabetes, carries on the multi dimensional analysis and the time series mining research, the main work is as follows:first, according to the characteristics of diabetes clinical data,the architecture design of the clinical data analysis and application system is given. The system includes data preprocessing module, multi dimension analysis module and time sequence mining module. Among them, the data preprocessing module mainly provides data support for system analysis. According to the system architecture, data preprocessing module will be mainly from two aspects: first, the data in HANA database, in accordance with the multi-dimensional analysis model, the multi-dimensional analysis of the fact table and dimension table; secondly, the clinical diagnostic data in accordance with the diagnostic event sequence, and obtain the patient’s data set.Then, with the specific characteristics of the clinical data of diabetes, the star model of the multi dimensional analysis of diabetes is constructed. Based on this, the SAP Business Object tool is designed and implemented. In the process of multi dimensional analysis, the resource allocation, the dynamic structure and data processing of SQL statements based on control and filter, and the two components Webi and Dashboards of SAP BO are used to realize the data of the diabetes, medication, diagnosis, and so on.Finally, in order to find out the regularity occurrence of diabetes’ complications,combined with the characteristics of diagnostic events, this paper carries on the event sequence of the historical clinical diabetes diagnostic data, and make a progress based on the traditional SPADE algorithm, and propose a new method based on time window for the clinical diagnosis of diabetes mellitus. The algorithm through time window setting, combined with the time interval of diabetes treatment, support for the time window of diabetes diagnosis frequent pattern discovery. Through experiments on real data sets, the validity and practical value of the proposed algorithm for the diabetes diagnosis based on time window is verified.