Knowledge gathered –
OPTIMAs Work Packages
Package Overview
To reach these objectives, the OPTIMA project is grouped into 9 work packages.
The Packages
Work Package 1
Project management
- Establishes a robust project management and governance structure to support all partners in achieving OPTIMA’s objectives.
- WP1’s responsibilities include technical and financial reporting, risk management processes and conflict resolution, critical path analysis for each WP, timely decision making and accountability.
- WP1 facilitates and optimises effective collaboration between all consortium partners, whilst ensuring all IMI2 requirements are respected. Key to this will be the development of an open and inclusive project environment that encourages all OPTIMA stakeholders to be actively involved.
Lead: European Association of Urology & Pfizer
Work Package 2
Informed consent, general requirement analysis, governance, and regulation
- Ensures an ethical and legal supporting framework that contributes to sharing data compliant to GDPR and building trust in AI solutions.
- Continuous ethical & legal advisory and monitoring of project’s development.
- Sets up an Ethics Advisory Board (EAB) to oversee these activities.
- Regulatory work towards CE marking of the developed software implementations of guidelines-based decision support tool will be performed, with the end-goal of providing a certifiable device.
Lead: University of Vienna & Hoffman-La Roche AG
Work Package 3
Guideline-based decision support tool
- Identifies the relevant guidelines for prostate, breast and lung cancer and translates these guidelines into electronic decision support tools, which will be integrated with EHR’s.
- Identifies critical research questions that can serve as input for the development of AI models and analyses.
- Captures retrospective and prospective data and integrates it into the data platform that will be used for building AI models.
- Assesses the impact of the support tools on guideline adherence.
Lead: Erasmus University Medical Center Rotterdam & Hoffman-La Roche AG
Work Package 4
Development Platform technical and infrastructural requirement analysis
- Gathers necessary technical and user requirements to develop an IT and knowledge platform to allow easy, compliant, and secured access to all the data collected or generated during the project to all members of the consortium. First focus will be on technical interoperability of existing systems. Open-source solutions will be used.
- Will work closely with WP5 regarding user personas and will focus on the general aspects.
- Will integrate the CIGs developed in WP3 and AI models in WP7 as part of the data platform. Develops a data management plan and will apply standards in data management to guarantee data security, data privacy and compliance with General Data Protection Regulation
- (GDPR) and ethical guidelines. Data will be made FAIR – Findable, accessible, interoperable, and reusable.
Lead: Information Technology for Translational Medicine & Amgen
Work Package 5
Platform implementation and evaluation
- Implement the platform in selected clinical sites for testing.
- Will use the technical work performed in WP4 to build, test and adapt the data platform for various types of users (e.g., clinicians, patients, data providers, bioinformaticians, sysadmins), to get access to the most relevant information and data visualisations.
- Pilots user champions in each of the user personas including clinical users to test the platform on a regular basis; the comments and requests for changes in the platform will then be submitted to WP4 as part of a continuous development cycle.
Lead: ASSOCIATION EISBM & Amgen
Work Package 6
(Non)interventional study data and real-world data gathering, preparation and integration
- Creates a data catalogue with patient-level data provided by beneficiaries and partner organisations from different countries. Through them, access to cancer patients, treatments, and outcomes relevant to existing guidelines and other prioritised research questions is enabled.
- Builds the data layer, that will provide a homogeneous and standardized way for the OPTIMA Platform to access data regardless of its origin, type, format, or location.
- Those data providers not uploading their data to the central OPTIMA servers will be nodes of a Federated Learning Network. This way the Machine Learning / Deep Learning models run in the OPTIMA platform will be able to be trained using datasets located in separated physical locations, without the need of sharing the data.
- In addition, we will work in close collaboration with WP4 and WP7 to create a centralised data catalogue for the analysis and mining of patient-level data to enable more complex AI analytics.
Lead: University of Oxford & Bayer AG
Work Package 7
Artificial Intelligence knowledge base implementation
- Lead: HZDR & Pfizer Will create explainable AI tools to support decisions in the clinics based on a variety of data sources and data sets provided by WP6 and informed by the gaps in guidelines prioritised by WP3. From the beginning we will include existing TRIPOD 3 or 4 AI models will be reused.
- A continuous machine-learning model deployment process of the AI models will allow permanent quality assurance of the models supported by WP5. Continuous updates and shipping of these updates to be implemented in WP4 as well as support of validity and impact assessment in WP3
- The platform will allow the development and integration of new AI models by future users subject to strict clinical validation (using relevant guidelines) and approval of the Scientific Governance Board.
Lead: CASUS, Helmholtz-Zentrum Dresden-Rossendorf & Pfizer
Work Package 8
Dissemination and communication
- Ensures broad communication and dissemination of OPTIMA’s results and outcomes to all stakeholders including patients, patient-organisations, regulatory agencies, healthcare providers, researchers, industry, healthcare payers, clinicians, policy makers and the general public.
- OPTIMA will leverage the communication networks for its diverse consortium partners to ensure tailored communication messages are developed and received by the project’s key audiences.
Lead: ARTTIC Innovation GmbH & Pfizer
Work Package 9
Exploitation and sustainability of the results
- Identify long-term sustainability strategies to ensure that the results and clinical decision support tools generated by OPTIMA are effectively implemented, built upon, and integrated into clinical practice benefiting patients into the future.
- Opportunities for valuable scientific collaboration are evaluated
- These ambitious goals will be achieved in collaboration with all stakeholders and will rely on the consortium building strong links with multiple organisations.
Lead: EAU & Pfizer
