Subsystem of prognostic risk factors analysis of childhood acute leukemias
Tom Ie, Ольга Владимировна Красько, Zalessky Ba, Novoselova Na, Suchkova Ap, Sotikova Ee, Skrygan Nm, Potapnev Mp, Uglova Ta · 2005
Propose: The purpose was to create the electronic register formed as a database where the results of medical examination for each patient are stored and to develop the software for statistical and intelligent data analysis applied to medical tasks of prediction outcome of induction therapy and patient risk group classification based on prognostic risk factor (PRF) combinations in childhood acute leukemias. Material and methods: Statistical and intelligent methods for data analysis, modern programming technologies for database and software development. The medical data were obtained in Belarussian Research Center for Pediatric Oncology and Hematology (BRCPOH). Results: The subsystem for multivariate statistical and intellectual analysis of data is realized as a special software “Professional Analysis of Prognostic Factors” (ProAPF). ProAPF includes a number of statistical and neural network analysis methods that were selected, upgraded and adopted according to the applied aim of the project. Conclusions: The database for prognostic risk factors (PRFs), collected at diagnosis and during early treatment course in children with acute lymphoblastic leukemia (ALL) and acute myeloid leukemia (AML), was created. The information technology of analysis of prognostic risk factors actualized as ProAPF software was developed to identify patient-specific risk group to apply risk group – oriented chemotherapy in patients with ALL or AML. The new PRFs and their combinations were explored, the combinations were selected to predict outcome of induction therapy.