WP 1 Project Management
Objectives: (a) To coordinate the scientific orientation of the project and to ensuring progression while maintaining its relevance, (b) To manage resources, monitor overall project performance and manage risks, ethics, and contingencies and (c) To establish appropriate communication channels with all participating project actors. Tasks: T1.1 – Coordination of the scientific orientation of the project, T1.2 – Resource and risk monitoring and management, T1.3 – Establishment of communication channels with all actors. Deliverables: D1.1 – Midterm Project Progress Report (M16), D1.2 – Project Final Report (M30).
WP 2 Use Cases, User Requirements, and Data Morphology
Objectives: (a) user requirements respecting all legal and ethics aspects (b) state of the art analysis (c) a technical and functional description of the architecture; (d) a complete description of NSCLC and CAD use cases (e) data organization and complete access of all diverse and heterogeneous participating data. Tasks: T2.1 – User requirements, T2.2 – State of the art analysis, T2.3 – EMERALD use cases, T2.4 – Morphology and organization of heterogeneous data. Deliverables: D2.1 – EMERALD User requirements analysis (M9), D2.2 – Survey paper (publication) (M9), D2.3 – EMERALD Use Cases and Experimental Scenarios (M10), D2.4 – EMERALD Database Repository (M12). Milestones: M1–EMERALD specs and repository completed (M12)
WP 3 New generation Χ-AI models using FCMs, ML & DL
Objectives: creation of new explainable AI models based on FCMs, ML and DL generating humanly understandable decisions and explicitly explaining such decisions. Tasks: T3.1–Theoretical dynamic advancements of the FCM model T3.2–DeepFCM: decision making model based on new dynamic FCMs and DL, T3.3–Explainability and trustworthiness assessment of XAI-based DeepFCMs. Deliverables: D3.1 – New Advancements in FCMs for Modeling Complexity and Decision Making (publication) (M14), D3.2 – DeepFCM model (publication) (M16), D3.3 – XAI-based DeepFCMs and Analysis of their explainability and trustworthiness (publication) (M18). Milestones: M2 – EMERALD new XAI models completed (M18).
WP 4 XAI - MDSS and Tools
Objectives: (a) advanced DM methodologies applied as multiple modules for primary types of data, (b) XAI prototype, incorporating dynamic FCMs and DeepFCMs produced in WP3, (c) open-source Python library with several algorithms implementing the XAI-based DeepFCMs and APIs and (d) a MDSS for the cases of NSCLC and CAD. Tasks: T4.1 – Advanced Data Mining, T4.2 – XAI prototype, T4.3 – Case study MDSS. Deliverables: D4.1 – Data Mining Tools for Medical Data (open source library) (M22), D4.2 – XAI Decision making report (at least one publication) (M25), D4.3 – Integrated MDSS prototype (at least one publication) (M25). Milestones: M3 – MDSS prototype and explainability completed (M25).
WP 5 MDSS Validation
Objectives: Evaluation of prognostic and diagnostic capabilities of the MDSS in T4.3 for both pilot cases by developing and validating models that can provide physicians with a survival probability for an individual NSCLC or CAD patient. EMERALD will assess the validity and reliability of the system in parallel on a predefined set of patients compared with real treatment scenarios to investigate the existence of prognostic and diagnostic variations. Tasks: T5.1 – EMERALD validation, T5.2 – Recommendations in relation to ethics, compliance with data protection. Del.: D5.1 – EMERALD MDSS validation report (2 publications) (M30), D5.2 – EMERALD ethics compliance report (M30). Milestones: M4 – Validation completed (M30).
WP 6 Dissemination routes
Objectives: EMERALD project will include a variety of activities ensuring the highest possible impact. The primary goal of WP6 is the wide spread of EMERALD outcomes among the appropriate target communities, at appropriate times, via appropriate methods. The dissemination and communication plan consists of the following tasks: Tasks: T6.1–website, T6.2–social networks, T6.3–related journal and conference publications, T6.4–Open-source and access library. Del.: D6.1–website (M3), D6.2 –social media (M1-M30), D6.3–Overview of publications (M7-M30), D6.4 – Python libraries available as open-source (M30). M: M5– Dissemination activities completed (M30), M6 - Completion of the open-source libraries (M30).
