The TRIACS analytical workflows platform for distributed clinical decision support

In this paper we discuss a flexible distributed workflow-based approach that enables researchers to study biomedical data for creating decision support pipelines. Specifically we describe the TRIACS platform, which has first been applied to supporting evidence-based decisions for optimum diabetic retinopathy screening intervals In the prioritization mechanism, pseudonymised case data is stratified for screening need by computation of outcome risk or by clustering of dasiaat-riskpsila cases with past cases of actual preventable outcomes. Workflows present a novel approach to this problem by providing an appropriate level of granularity for breaking the domain problem into a collection of reusable service-oriented components that can be applied in different ways. TRIACS is intended to make the creation of new application logic quicker and easier than bespoke development methods. Through the TRIACS workflow interface, modular code is portable and available to solve analogous domain problems including application to trial studies for mining and analysing clinical data.