Heterogeneous Data Integration in Healthcare Big Data: A Comprehensive Systematic Review

Healthcare data integration is one of the most complex challenges in modern medical informatics. Hospitals and research centers collect huge amounts of data from imaging, electronic health records, genomics, proteomics, and sensor devices. Some repositories exceed three petabytes of information [1]. These data arrive in different formats and at different times. Researchers need methods that can combine such diverse information while keeping it useful for doctors and patients. In this review we summarize current methods for integrating multimodal healthcare data. We focus on how these methods scale, how they are evaluated, and how they can be used in real clinics. We also highlight the importance of integrating data across different health domains such as oncology, neurology, mental health and lifestyle studies [54][61].