UBITECH is hosting the kick-off meeting, in Athens, Greece (January 28-29, 2020), of the ASCAPE Research and Innovation Action, officially started on January 1st, 2020. The project is funded by European Commission under Horizon 2020 Programme (Grant Agreement No. 875351) and spans on the period January 2020 – December 2022. The ASCAPE project aims to take advantage of the recent ICT advances in Big Data, Artificial Intelligence and Machine Learning to support cancer patients’ quality of life and health status. To achieve its objective, ASCAPE will create an open AI infrastructure that will enable health stakeholders (hospitals, research institutions, companies, etc.) to deploy an edge-node locally and execute the AI algorithmson their private data without the need to send their data outside their domain. Any new knowledge produced by this process will be sent back to the open AI infrastructure. This way the knowledge will be shared among everyone and the medical data will remain private. With this feature, ASCAPE is aiming at democratizing access to cancer support. In particular, with ASCAPE even less developed countries or small healthcare providers (small rural hospitals) that do not possess large amounts of medical data, can connect to ASCAPE and benefit from its collective knowledge.
Among ASCAPE goals is for the AI infrastructure to continue growing long after the completion of the project, building up knowledge for many types of cancer. The long-term vision is to create a framework where multiple healthcare stakeholders will participate to build and share knowledge about various types of cancer and if possible other kinds of health issues. For the duration of this project however, ASCAPE will focus the training of this AI in two types of cancer: breast and prostate. The reasons for this decision are a) to achieve sufficient coverage across genders and age groups, hence facilitating its ongoing improvements and applicability towards any type of cancer in the future and b) to achieve high impact since breast and prostate cancer have a high survival rate and patients that live longer and will benefit more from our framework. To train the proposed AI infrastructure ASCAPE will collect different types of datasets from multiple sources: In particular, we will collect a) large datasets from the project’s healthcare partners, open online sources, institutions etc.; b) datasets collected through active monitoring of recruited patients (from project’s pilots) and c) datasets from organisations carefully selected in our open call that will take place during the final year of the project.
Within ASCAPE, UBITECH has undertaken the administrative and financial project management, addressing the ethical requirements, as well. Moreover, UBITECH R&D team will contribute towards the implementation of the ASCAPE Open AI infrastructure that aggregates data from various data sources and edge nodes and possesses computational and storage capabilities for scalable processing of these data. Finally, UBITECH will work towards the design and development of the edge-nodes – to be deployed locally by external parties, representing resources that possess computational, storage and communication capabilities and participate in the process of monitoring, learning and predicting the patients’ health status. The collection of Edge-nodes defines the Federated Learning subsystem which is responsible to build and sustain a global Federated Deep Learning model.