European in silico medicine infrastructure established: VPH-Share business plan successfully presented

08/26/2015

In silico medicine, the use of computer simulation in the diagnosis, treatment, and prevention of a disease, compliments in vivo and in vitro medicine. In silico medicine is a synergy of medical informatics, biocomputational modelling, and bioinformatics. Its potentially high transformational impact on healthcare has been promoted by the European Commission since 2001. The European Virtual Physiological Human (VPH) Initiative has spawned collaborative projects across disciplines and with international partners in industry, academia, and healthcare.

To establish the urgently needed infrastructure for this research, beginning in 2011, the European Commission has funded within its FP7 Programme the large-scale integrated project VPH-Share - Virtual Physiological Human: Sharing for Healthcare. This project has developed the tools and services to allow the European and global VPH community to share data, tools and knowledge for multiscale models, to design new disease-specific VPH workflows, and to identify approaches for clinical implementation and business case development.

The final technical review of the project by EC experts took place on 2-3 July, 2015, in Eindhoven, The Netherlands. It was hosted jointly between Catharina Hospital  and Philips , with TU/e hosting the rehearsals. During the review four user stories involving multiple demonstrations were presented, demonstrating the maturity of the infrastructure realised by this project. Prof. Dr. Karl Stroetmann participated on behalf of empirica, presenting the final VPH-Share business plan, which the company developed as leader of the exploitation and evaluation work package, in close cooperation with all other project partners.

The four user stories concerned:

The Clinical User Story demonstrated the importance of modelling a patient’s Fractional Flow Reserve. Through access to cloud and high performance (HPC) computing services VPH-Share enables the comprehensive sensitivity analysis of models, and simulations over the physiological envelope of patient parameters as well as multiple runs of computationally expensive models.

The Educational User Story exploited the scalability provided by VPH-Share to provide access to a virtual patient educational model, which runs in Open Labyrinth and was delivered in a MOOC (Massive Open Online Course), providing access to over 20,000 learners.

The Industrial User Story, making use of the Share infrastructure and tools, was presented by a scientist of Philips Research BV, Eindhoven. It used machine learning to predict the readmission risks of Acute Coronary Syndrome; showing how the morphology of aneurisms can be clustered into six distinct groups; and finally demonstrated the use of Azure’s Machine Learning interface on a publicly available dataset. A representative of Microsoft Research Connections underscored the importance of initiatives such as VPH-Share in providing Research as a Service.

The Research User Story demonstrated a recently added functionality of Share which allows any third party project to brand and promote its own infrastructure as if it were a stand-alone portal, through which it can make visible and available the respective tools, models, data and workflows relevant to that project group as well as the wider community.

Throughout the review, the project management set the context and wove together the strands of the review, emphasising the exploitation of VPH-Share by different user groups and how the sustainability of VPH-Share has been pursued through interactions with external VPH stakeholders, including cloud and software providers and supporting the four types of users illustrated by the user stories.

The project was enthusiastically received by reviewers and received a rare “excellent” rating. It met all its objectives and exceeded the work outlined in its Description of Work. Particular mention was made of the positive, collaborative and open attitude of the project and its services to users.

More information on the project can be found here.