CBmed & SAP: Mining Highly Diverse Biomedical Data for Biomarker Research
Colleagues from CBmed visited the SAP offices in Walldorf, and I had a chance to speak with the CEOs, Robert Fasching (CFO) and Thomas Pieber (CSO) to learn more about CBMed’s activities.
Can you tell me more about CBmed?
CBmed’s objective is to become a global leading center for biomarker research, especially in the areas of cancer, metabolism and inflammation. Biomarkers are biologically measurable units (e.g. enzymes, hormones or genes) that make early detection and individual therapy of diseases possible. The “Center For Biomarker Research In Medicine” (CBmed GmbH, Graz, Austria) links the scientific expertise of, for example, the medical universities in Graz and Vienna with innovative technologies from leading pharmaceutical, diagnostic and medical technology companies. We thus work closely with the many stakeholders in the area of Precision Medicine. One of our main objectives is the assimilation and mining of data. Our core labs use innovative technologies such as Next Generation Sequencing, mass-spectrometry-based metabolomics, digital pathology and proteomics.
What does your cooperation with SAP look like?
One focus of CBmed is data integration and information discovery; biomarker innovation and validation is only possible if vast amounts of heterogeneous, multi-dimensional, structured and un-structured data sets are combined. We have evaluated several technology providers in this area, and ultimately decided to use the SAP Connected Health platform (based on SAP HANA) as a central part of our knowledge system. SAP Connected Health platform allows us to integrate the different data sets and to analyze the data in real-time.
What will the CBmed system look like in detail?
The system is currently being build. The project data will come from different systems that contain clinical routine documentation, and are automatically standardized to make them available for innovative applications. The core of the system will be a high-performance in-memory database using SAP HANA and a “virtual biomedical warehouse” created through SAP Connected Health platform. Since an important part of the clinical information is available only as text, it has to be analyzed by means of speech-based techniques and mapped to a standardized vocabulary. The data obtained in this manner will be combined with other information, such as laboratory parameters, diagnostic codes, etc. and will enable semantically standardized patient profiles. This will ultimately speed up research projects – but will also lead to clinical decision support for the individual physician in the hospital. For this purpose, innovative applications will be created that allow quick and easy interpretation of the information. These will include e.g. the four initial concept apps Patient QuickView, Coding, Recruiting and Prediction, each supported by an underlying workflow and a presentation layer for the end-user. Patient QuickView will provide an automatic summary of decision-relevant patient data, depending on the preferences of the respective user. Coding assists physicians in the coding of cases for administrative purposes and submitting suggestions for appropriate diagnosis and procedure codes. Recruiting will make it easier to define patient groups for specific research projects, for example in biomarker research. Prediction will help predict future events, such as resumptions, based on semantically enriched patient profiles.
We are also already thinking about other applications. We have access to an almost endless range of data, the evaluation of which will be key for the development of personalized medicine. The exciting thing: use of this innovative platform provides an alternative to time-consuming research in electronic patient records. We will finally be able to access uniform, standardized data and to significantly optimize workflows in the clinical and research environment, and thus relieve the persons involved.
Join SAP at HIMSS 2017 in Orlando to learn more about this joint project: booth #543.