DQ0 Use Case Healthcare
Make sensitive clinical data available for advanced analytics. The data can be used by external experts without ever leaving the clinic. No need to request ethics approval.
DQ0 Solutions for Healthcare
DQ0 removes all barriers
when working with
sensitive patient data
Analyzing medical records and patient data with machine learning techniques has huge potential for better diagnostics or treatment recommendations.
But patient data is highly sensitive. The clinic’s data privacy officer must ensure that sensitive information never leaves the clinic when working with the data. Hence access to valuable data sources is often very restricted or even impossible.
Moreover, If a way is found to work with the data, it often involves complex and error-prone processing steps to protect the data. De-identification techniques like anonymization are implemented with scripts that transform a sensitive database to an allegedly more secure second database for analysis that only contains a subset of the original data.
These scripts are labor intensive and must be adjusted constantly. In addition, the results are not secure. Anonymous patient data often still allows conclusions to be drawn about secret information, e.g. patient identification.
DQ0 protects the patient data right where it is. Without any modifications and without any direct access.
With DQ0 data analysis is performed in the secure data enclave, inside the clinic. The clinic’s data privacy officer remains in full control over the data and everything that happens with it.
Data scientists work with the data through DQ0’s secure interface. As privacy is baked directly into DQ0, granting access to outside data scientists is no longer a question of data protection.
Every result that data scientists receive via DQ0 is protected through mathematically proven privacy principles. Working with DQ0-provided data sets is like working with perfectly general, PII-free versions of it.
DQ0 protects data sets at computation level. There is no need for data de-identification processing steps. Data sets can be made available wholly and directly.