Probabilistic methods for multi-source and temporal biomedical data quality assessment

Speaker: Juan M Garcia-Gomez

Institution: Biomedical Data Mining Group, Universitat Politecnica de Valencia, Spain.

Dr Juan Miguel Garcia-Gomez is a Computer Scientist and Senior Lecturer at Universitat Politècnica de València (UPV), Spain. He leads the UPV’s Biomedical Data Science Lab, and his research focuses in machine learning and data-driven approaches for improving healthcare, including clinical decision support systems for diagnosis and treatment of brain tumours and breast cancer, and also automatic biomedical data quality assessment tools for big data repositories. He has authored more than 40 articles on specialized journals and in 2014 his work titled An HL7-CDA wrapper for facilitating semantic interoperability to rule-based Clinical Decision Support Systems (doi:10.1016/j.cmpb.2012.10.003) was selected as best Health Information Systems paper by the International Medical Informatics Association. He coordinated the pattern recognition work for the EU FP6 eTUMOUR project (LSHC-CT-2004;503094), the predictive analysis work for the EU FP6 HealthAgents Project (IST-2004;27214), and the decision support work for the EU FP7  Help4Mood (ICT-2009-4;248765). 

Abstract: Nowadays, biomedical research and decision making depend to a great extent on the data stored in information systems. As a consequence, a lack of data quality (DQ) may have significant effects in the interpretation of data, which may lead to suboptimal decisions, or hinder the derived research processes and outcomes. This talk is about the research and development of methods for assessing two DQ problems of special importance in large-scale multi-site repositories acquired during long periods of time: (1) the variability of data probability distributions among different data sources or sites, and (2) the variability of data probability distributions over time. 

After a short introduction to his Biomedical Data Science Lab, Dr. Juan M Garcia-Gomez will review and show the application of new metrics based on information theory and information geometry for assessing spatio-temporal variability for multi-type, multivariate and multi-modal data.