Innovative benchmarking approaches to improve COmputational METHods for big data analysis in Health

If analysed correctly, real health data can become a valuable tool. COMETH develops a framework for assessing benchmarks to analyse health data and then trains doctors to use that framework.


COMETH targets the need to evaluate novel computational methods and apply them directly on real health data sets, while also offering training to help professionals meet this need. The programme will produce a benchmarking framework that can be used with the data that health data scientists and health professionals handle every day. As a proof of concept, the programme will apply the framework to quantification of cell heterogeneity in cancer, an area of study that can guide creation of improved cancer treatment strategies. But COMETH’s benchmarking framework will also lend itself to a variety of other applications.

The COMETH programme will provide a digital platform containing statistical methods to be benchmarked and data sets to be developed by health data scientists, as well as a web application dedicated to the direct use of these tools on novel clinical data sets provided by health professionals. COMETH will also organise a two-day course for clinicians and cancer pathologists on the application of these statistical methods.

Magali Richard
| Activity Lead | UGA