Solução de auxílio ao diagnóstico e à pesquisa médica baseada em mineração de dados utilizando interface android

Abstract

Since the primary studies on the applications of Information Technology aiming to add value to other areas of knowledge, the playing eld of medicine has always been seen as fertile ground for such. With the advent of Arti cial Intelligence techniques, computer programs have been given a power of learning more sophisticated and thus opening the possibility of its use beyond the hospital administrative processes, drawing ever closer to the provision of patient care. Therefore, this paper proposes to demonstrate the feasibility of an aid to medical diagnosis and obtaining implicit knowledge in databases of three diseases: breast cancer, dermatology and vertebral column problems. To do so, is applied the process of extracting knowledge from databases in order to achieve these goals. This process has Data Mining as its core, which in turn relies on machine learning algorithms to transform data, sometimes not analyzed, in useful information for business referred to, in this case about health care. Therefore, this work presents a study aided by the tool Weka, to determine which machine learning algorithms perform best when applied to target databases. With these algorithms in hand, is implemented a solution to aid the diagnosis and study of medical applications making use of Android as interface for healthcare professionals, with it, utilizing what is most modern in terms of mobile electronic devices in the world market. The results were quite satisfactory, given that the objectives for the study on the determination of Data Mining algorithms, preparation of databases for future research and implementation of the solution for the diagnosis have been met and, together, prove that you can apply tools of information technology to add value to medical practice.