A scientific paper entitled “Utilizing High Performance Computing Techniques for Efficiently Anonymizing Sensitive Patient Data” has been co-authored by UBITECH and published at the Collaborative European Research Conference CERC 2017 that is held in Karlsruhe, Germany at September 22-23, 2017. In this paper, UBITECH presents how High Performance Computing (HPC) enhances the adoption of Data Cubes technology for the anonymization of large sets of patient data in order to be able to cope with very large data sets. Indeed, in the case of genetic data, it may be possible that the size of original data may become very big, since the gene expression information or Single Nucleotide Polymorphisms (SNPs) that need to be recorded per patient may be in the order of hundreds of thousands.
In particular, using HPC has been demonstrated to offer a significant speedup during the generation of anonymized patient data. This facilitates the real time request and acquisition of data instead of having to wait or having to pre-generating data that could become outdated. Future work consists of evaluating the response times of the algorithm to various combinations of sample test data as well as monitoring and measuring processor idle times. The latter will help to determine more accurately what are the bottlenecks that introduce the biggest deviations from the theoretical minimal times and it will provide hints to further enhance the efficiency of the parallelization schema.
Source: http://www.cerc-conference.eu/wp-content/uploads/2015/09/proceedingscerc2017.pdf
