Knowledge visualization for supporting communication in cardiovascular risk assessment hypotheses

Abstract

Cardiovascular diseases (CVD) represent a severe cause in mortality cases over the world. Early detection and foreseeing of future developments are essential goals for every cardiologist. For this reason, detecting and understanding several factors that trigger cardiovascular diseases is essential. As we have emphasized in eProcord spreadsheets (a 3-year research project that assessed the cardiovascular risk through medical and modeling methods and identified patterns for patients), Database Management Systems, statistical, and Machine-Learning soft-wares are tools which can be used in order to identify these factors through data analysis. This paper presents an approach towards concept data analysis of medical data using a mathematization of the classical notion of concept which has been implemented in the knowledge management suite ToscanaJ, grounded on the Conceptual Knowledge Processing paradigm. Starting from this premise, we propose a human centered method to investigate, represent, process and acquire knowledge from a medical database. This method offers a reasoning support for an expert centered visualization of previous collected data which facilitates a better understanding of cardiovascular risk assessment hypotheses, in order to ground a solid environment for these hypotheses. We use the visualization capabilities of Formal Concept Analysis in order to explore the knowledge encoded in these data and we give an overview on how the Conceptual Knowledge Processing methods can be used as a knowledge discovery tool for datasets related to cardiovascular diseases.