Publishing Electronic Medical Record akin to Privacy

Advancement in medical technology has resulted in offering sophisticated treatments with the help of medical databases. Medical database includes historic and progressive medical records of patients stored in medical centers and hospitals. Electronic Medical Record (EMR) managed and stored electronically provides extensive information for medical learning and research analysis. Physicians often treat medical cases and determine treatments referring similar EMR from the medical repository. In general the EMRs are restricted within the hospital zones. It is often required to share EMR data with data analyst who extracts useful information for social welfare reasons. In such conditions privacy information of the patients need to be preserved while retaining data utility. When data utility is compromised, then the data analysis will lead to wrong outcomes. However there are methods to sanitize EMR data such that privacy and accuracy are preserved. Conventional methods have constraints in managing privacy and accuracy together. This paper explains a new technique called Generalized Average Partition (GAP) which is designed exclusively for EMR data handling. Privacy and accuracy losses are calculated for selected conventional methods to demonstrate the level of accuracy and privacy losses.