Skip to the content Skip to the NavigationAnalyzing Electronic Medical Records Data
- Using data mining technology to extract and recommend the optimal flow of medical treatment from the history of medical treatment in large amounts of electronic medical records.
- Providing the tacit knowledge of experienced doctors as a pattern of medical practice and estimating factors in the pattern
- By clarifying commonalities and differences in patterns between multiple medical institutions, it is possible to confirm the characteristics of one’s own medical institution and improve medical practices by referring to treatment patterns of other medical institutions
Guaranteeing Data Privacy During Analysis
- Local/global differential privacy, which adds noise to raw data, can greatly contribute to data analysis that does not identify individuals.
- Aiming to develop more accurate and faster data protection analysis
Power-proportioned Distributed File Systems
- In the Hadoop Distributed File System (Hadoop Distributed File System), which is widely used for processing big data, it provides performance according to the power consumption of the operating nodes
- Employ a data placement using data replication distributes the load among active nodes
- Prevent performance degradation when changing system configuration by devising data placement on nodes that can be stopped
- Further improve the performance by a distributed matadata management
