From quite some time, biologists have been gathering big amounts of biomarker data from patients suffering specific illnesses and from healthy people. Their problem now lies in the processing of that huge amount of data that will enable them extracting meaningful information about the links and thus the rules enabling a diagnosis based on specific biomarkers. In this paper we propose an approach to this problem using fuzzy logic to model the diagnostic systems and evolutionary computing to find such systems. Moreover, the speed of execution of the proposed design which is based on several Virtex5 FPGAs with respect to a standard software computation, enables the realization of thousands of successive evolutionary runs within a reasonable time and thus permits to obtain robust statistical information enabling the selection of meaningful biomarkers for the diagnosis of specific diseases.