
AVATAR WORK PACKAGES
The scientific and technological objectives of the project are:
STO1: Algorithm flowchart. The algorithm will be based on AI techniques and ML models. STO1 is in a way, the exact problem definition for the project and will guide the work for the next nine months. This step ensures that the data parameters used for the testing, the workflow of the algorithm and the validation steps will be captured and documented correctly from the beginning of the project. The output from this STO1 will be a very detailed flowchart that guide the technical developments (STO2 and STO3).
STO2: Procedure to adapt the methodology for cancer data. Real data from patients with prostate cancer will be used and fed into the Deep Generative Models (DGMs) to generate synthetic data. The real patient data will be first cleansed using data quality routines (already developed by team members scientists) for data validity and sanity. The filtered (cleansed) data will be then split into a train file and a test file. Different supervised models will be implemented using the split train and test subsets. The train file will be used for the training procedure, while the test file will be used for evaluating the performance of the models. The final synthetic data generator model will be created from an ensemble of models.
STO3: Evaluation. The final model will generate the desired synthetic data by the completion of STO2. In STO3, the quality of the generated synthetic data will be assessed. During the evaluation procedure, different utility assessment methods for synthetic data and privacy metrics will be used to define the extent to which the statistical properties of the real data are captured to the synthetic data sets and how much of the real data may be revealed (directly or indirectly) by the synthetic data.
​
The STOs and impact of the project are achieved through the WPs of the project.
​
WP1, PROJECT MANAGEMENT
This work package (WP) oversees the project. The main objectives are to: a) communicate decisions, results, coordinate meetings and tasks among the team members and ensure that there is adequate collaboration between them, b) to monitor the work during the project in order to meet project milestones and objectives, c) to identify and mitigate risks that could affect the project’s deliverables or its overall impact and d) generate all technical and financial reports for the funding agency (i.e., RIF).
​
WP2, DISSEMINATION AND EXPLOITATION ACTIVITIES
The objectives of WP2 are to disseminate the knowledge, results, methodology and new techniques to scientists and engineers in the field of digital health. WP2 also informs the public and other stakeholders and innovators as a whole to generate interest in the project activities and results. Another important objective of WP2 is to secure subsequent private funding to: a) ensure the continuation of this project (to reach a higher TRL) and b) support the sustainable and continued collaboration between the partners.
​
WP3, SYNTHETIC DATA ALGORITHM DEVELOPMENT AND SECURING OF REAL HEALTHCARE DATA FOR VALIDATION
The main objectives of WP3 are to fulfil STO1 and STO2 and include the definition of the workflow for generating synthetic data, the collection of real patient data and the implementation of the synthetic generator model using the collected data.
​
WP4, VALIDATION OF THE PRIVACY AND UTILITY OF THE SYNTHETIC DATA
The main objectives of WP4 are to fulfil STO3 and include the formulation of the evaluation metrics list, and the validation of the results obtained from the synthetic data against the original data.
​
​
​
