COPDGene Omics Data Warehouse and Data Mining
Since January 2017
This project, which is sponsored by the National Institutes of Health (NIH), is two-fold. For the first component of this project, we developed tools for efficient data integration, storage, query, and analysis for the effective data-driven study of COPD. As for the second component, we are currently applying various high-dimensional clustering methods, such as subspace and projective clustering algorithms, to the COPDGene datasets. For more information, please visit the BDLab website.
● Technologies: C#, WPF, LINQ, SQL, R, Python, sckit-learn
Synthea-OMOP ETL Project
This ETL application extracts data from Synthea FHIR outputs, transforms it to the OMOP Common Data Model, and loads it to an OMOP-based database hosted on PostgreSQL server. This application also handles missing data, missing vocabularies, etc.
● Roles: design, implementation, and testing the application using agile methodology
● Technologies: Java, JSON, PostgreSQL, IntelliJ
As a part of this undergoing project, we are developing a suite of novel tools for efficient data integration, storage, querying, and analysis, all as part of an end-to-end data management and mining system that enables effective data-driven study of COPD. As an example, this tool is going to be used as an underlying layer for applying machine learning techniques on COPDGene data. For more info please visit BDLab website.
● Technologies: C#, WPF, Microsoft SQL Server, LINQ, Microsoft Visual Studio
The RadOn project focuses on developing data integration and management solutions to automate verification of radiation oncology treatments. In particular, such solutions are expected to compare the prescription provided by a physician, treatment plans created by dosimetrists, and the details of actual treatment carried out by the treatment instruments, in order to identify inconsistencies in the treatment accurately. For more info please visit BDLab website.
● Technologies: C#, WPF, Microsoft SQL Server, LINQ, Eclipse Scripting API
Implemented an application using Microsoft StreamInsight (StreamInsight) to receive streaming sensor data from Chevron, detect errors, and reconstruct them.
This project was intended for USC Department of Public Safety to detect and announce crimes based on Geospatial data. The final application was implemented using Oracle Spatial Database, Microsoft SQL Server, Java, JDBC, and Microsoft C#.NET