Requirements Federated Learning and mUlti-party computation Techniques for prostatE cancer
0.1.0 - ci-build
Funded by the European Union

Requirements Federated Learning and mUlti-party computation Techniques for prostatE cancer, published by HL7 Europe. This guide is not an authorized publication; it is the continuous build for version 0.1.0 built by the FHIR (HL7® FHIR® Standard) CI Build. This version is based on the current content of https://github.com/hl7-eu/flute-requirements/ and changes regularly. See the Directory of published versions

4.4 Implications on architecture

Table 7: Human interface and user-facing features - functional requirements

ID Description Priority Category
F-HUF-1 Researcher node has jupyterlab interface High user interface
F-HUF-2 Researcher node offers all features provided by TRUMPET researcher node Medium algorithm
F-HUF-3 Support for federated Logistic Regression (LR) High algorithm
F-HUF-4 Support for federated Decision Trees (DT) High algorithm
F-HUF-5 Support for federated Random Forests (RF) Medium algorithm
F-HUF-6 Support for federated Support Vector Machines (SVM) Low algorithm
F-HUF-7 Support for federated Deep Neural Networks (DNN) High algorithm
F-HUF-8 Support for federated Convolutional Neural Networks (CNN) Medium algorithm
F-HUF-9 Support for federated Bayesian Optimization Medium algorithm
F-HUF-10 Support for federated Grid Optimization High algorithm
F-HUF-11 Support for federated Generative Adversarial Networks (GAN) Medium algorithm
F-HUF-12 Support for federated Variational auto-encoders (VAE) Medium algorithm
F-HUF-13 Support for federated Diffusion models Medium algorithm
F-HUF-14 Support for at least one effective federated synthetic data generator learner (GAN, VAE, or DiffMod) High algorithm
F-HUF-15 Support for multi-model synthetic health data (both tabular & image) High algorithm
F-HUF-16 Synthetic data generation module should allow for specifying what data (images, tabular …) should be generated High user interface
F-HUF-17 Synthetic data generation module should allow for specifying population subsets, e.g., only with cancer High user interface
F-HUF-18 Generation of synthetic 3D MRI images High algorithm
F-HUF-19 Data owner node has functional interface with local data owner database High integration
F-HUF-20 Data owner node has user interface for data owner users High user interface
F-HUF-21 Data owner node has server interfacing with other nodes High integration

Table 8: Human interface and user-facing features - non-functional requirements

ID Description Priority Category
NF-HUF-1 All output preserves privacy High privacy
NF-HUF-2 All algorithm implementations should follow the platform guidelines (adopted & revised from TRUMPET), e.g., on privacy/security parameters High privacy