hasantuberlin / impro-project

Integration of heterogeneous data and analytics in the Medical domain

Goal: One of the difficulties encountered in healthcare analytics nowadays, apart from the huge generation and availability of data, is the heterogeneity of the data [1,2]. In a typical healthcare analytics use case very different types of structured, semi-structured and not structured at all data should be considered. Examples include Electronic Medi- cal Records (EMR), Electronic Health Records (EHR), images, medical text reports, time series, etc. The objective of this project is to study how these heterogeneous data can be integrated into a unified model and system that allows scalable data analytics and exploration of information. Such a model should facilitate visualization, via a dashboard, of historical events and signals (time series measurements such as ECG or blood pressure) collected for many patients. The objective is to make this project as practical as possible, therefore the scenario and data available in one of the PhysioNet academic challenges in Healthcare will be used: Predicting Acute Hypotensive Episodes, A challenge from PhysioNet and Computers in Cardiology 2009 [3]. The solution should be scalable so modern distribute data flow sys- tems, such as Apache Flink, should be considered for processing, as well a time series data base or flexible NoSQL databases for storing.

Data from PhysioNet/Cinc Challenge 2009 https://archive.physionet.org/challenge/2009/

References:

1. J. Andreu-Perez, C. C. Y. Poon, R. D. Merrifield, S. T. C. Wong, and G. Z. Yang, “Big Data for Health,” IEEE Journal of Biomedical and Health Informatics, vol. 19, no. 4, pp. 1193–1208, Jul.
3. A. R. Reddy and P. S. Kumar, “Predictive Big Data Analytics in Healthcare,” in 2016 Second Inter- national Conference on Computational Intelligence Communication Technology (CICT), 2016, pp. 623–626.
4. GB Moody, LH Lehman, “Predicting acute hypotensive episodes: The 10th Annual Physio- Net/Computers in Cardiology Challenge”, https://physionet.org/challenge/2009/papers/0541.pdfbd