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

5.3 Cross-border prostate cancer pilot study Requirements

Requirements for the AI models FLUTE should consider impacts on the use of developed AI models, in particular when designing, developing and using artificial intelligence, FLUTE should assess the impact of this artificial intelligence on human rights, particularly the right to the protection of personal data, as well as human dignity and other fundamental values. This assessment should be part of a broader data protection impact assessment and should take into account the requirements set by the High-Level Expert Group on Artificial Intelligence. Above requirements should be continuously evaluated and addressed throughout the AI system’s lifecycle. The project should observe the developments regarding draft AI Act and - where appropriate - align with the standards for AI development that are set therein. The requirements will be in more detail described in T6.1: Legal and ethical compliance of the FLUTE platform.

Variables used in the previous models Predictive variables for BCN-RC 1

1. Age at the biopsy (years without decimals)
2. PCa family history (no =0 vs. yes =1)
3. Type of biopsy (initial =0 vs. repeated =2)
4. PSA (ng/ml with one decimal)
5. DRE (normal =0 vs. suspicious =1)
6. DRE-prostate volume category (small =1 vs. median =2 vs. large =3) Outcome variable for BCN-RC 1 ISUP-GG (0 =no PCa; 1 =Gleason 3+3; 2 = Gleason 3+4; 3 =Gleason 4+3; 4 =Gleason 4+4; 5 = Gleason 4+5, 5+4, or 5+5)

Predictive variables for the BCN-RC 2

1. Age at the biopsy (years without decimals)
2. PCa family history (no =0 vs. yes =1)
3. Type of biopsy (initial =0 vs. repeated =2)
4. PSA (ng/ml with one decimal)
5. DRE (normal =0 vs. suspicious =1)
6. MRI-prostate volume (cc with one decimal)
7. PI-RADS v.2.0 or 2.1 (1 to 5) Outcome variable for BCN-RC 2 ISUP-GG (0 =no PCa; 1 =Gleason 3+3; 2 = Gleason 3+4; 3 =Gleason 4+3; 4 =Gleason 4+4; 5 = Gleason 4+5, 5+4, or 5+5)

Table 9: Cross-border prostate cancer pilot study – functional requirements

ID	Description	Priority	Category
F-PIL-1	mpMRI/bpMRI shall be performed within 1 year prior to the prostate biopsy		Cohort inclusion criteria
F-PIL-2	Cohorts shall consist of men with clinical suspicion of PCa based on a PSA > 3.0 ng/ml and/or abnormal DRE		Cohort inclusion criteria
F-PIL-3	Lesions detected in mpMRI/bpMRI shall have to be reported using the Prostate Imaging-Report and Data System (PI-RADS) in version 2.0 or higher.		Cohort inclusion criteria
F-PIL-4	Prostate biopsies shall be systematic and targeted in cases of PI-RADS ≥3 lesions.		Cohort inclusion criteria
F-PIL-5	The platform shall define the methodology to extract/load/transform data from clinical databases and data warehouse into the FLUTE data node.		Platform
F-PIL-6	Input data shall be anonymized or pseudonymized		Clinical data
F-PIL-7	Clinical data and MRI (both raw and processed) shall be linked		Data inputs
F-PIL-8	MRI imaging study shall comply with specific requirements of QP-Prostate tool provided for FLUTE project.		MRI images
F-PIL-9	Data shall be labelled with class csPCa 0 or 1		Data inputs
F-PIL-10	Reduced sample datasets with the 7 clinical variables and associated images shall be shared to generate synthetic images and algorithms		Platform
F-PIL-11	No sensitive information shall be revealed from exchanged messages (aggregates, models statistics, etc.) between users		Platform
F-PIL-12	Access to local FLUTE nodes shall be Controlled/restricted		Platform
F-PIL-13	Platform shall allow AI developers to train their models in accordance with their legal requirements and document such training		Platform(s)
F-PIL-14	The training requests sent to the FLUTE nodes shall specify the minimum/maximum resources needed to be executed.		Platform
F-PIL-15	The performance of the model shall be higher than the BCN1 and BCN2 models		Algorithm
F-PIL-16	AI researchers shall be able to discover and select the datasets registered at each FLUTE node and obtain descriptive statistics about the datasets		Platform
F-PIL-17	Jupyter notebooks shall be integrated with the FLUTE functionalities to ensure discoverability of the datasets.		Platform
F-PIL-18	Platform shall allow the AI researchers to search for the relevant dataset.		Platform
F-PIL-19	Platform shall provide space to add guidance documents and instructions on how to use the Platform and the datasets.		Platform
F-PIL-20	Platform shall allow authentication of authorized individuals from Data owners and Data Users and varied level of access, based on their defined roles		Platform
F-PIL-21	Platform shall keep record of Data Users and Data owners and logs details of their activity in the Platform		Platform
F-PIL-22	Platform shall ensure that the training data remains on the federated node and any processing, analysis and AI training is performed there. Data User shall not see, directly access or download the data, i.e. the AI model shall only be trained in the local node		AI training
F-PIL-23	There shall be a security check of the uploaded AI model prior to its deployment in the FLUTE data.		Platform
F-PIL-24	AI models BCN1/BCN2 trained through the platform shall be packaged into software components and deployed at the clinical sites involved in validation activities.		Pilot
F-PIL-25	The platform SHALL be able to generate synthetic data for mpMRI, bpMRI and tabular data for BCN1 and BCN2 case series.		Pilot
F-PIL-26	The platform SHALL be able to train BCN1 and BCN2 models from an augmented/balanced datasets thanks to synthetic data.		AI training

Table 10: Cross-border prostate cancer pilot study – non-functional requirements

ID	Description	Priority	Category
NF-PIL-1	Units shall be harmonized		Clinical data
NF-PIL-2	A common (FHIR) data model shall be defined to represent the clinical data used in the study.		Platform
NF-PIL-3	The platform shall provide validators that check whether the clinical data pushed to the local node complies with the common data model.		Platform
NF-PIL-4	The data shall be standardized to a common (FHIR) data model before ingestion into the FLUTE local node.		Clinical data.
NF-PIL-5	Cryptographic methods like homomorphic encryption and differential privacy shall be used to aggregate statistics about the cohort without disclosing (leaking) sensitive data outside the local node		Security
NF-PIL-6	Platform shall keep and display FLUTE data catalogue with defined basic metadata that characterizes the datasets available through the FLUTE Platform.		Legal/Data management Platform
NF-PIL-7	Platform shall display terms and conditions of use (T&C) and Privacy policy.		Legal
NF-PIL-8	Platform shall display the conditions of the use of each of the datasets, as specified by its owner or the data hub.		Legal
NF-PIL-9	Datasets which are not defined as open to all users of the platform, shall only be available to uses which request access to the dataset and are permitted to use it by the data owner or data hub.		Legal
NF-PIL-10	Platform shall monitor the use of the data in the Platform, to detect potential misuse. It shall implement measures for detection of data breaches and potential privacy threats/leaks		Security
NF-PIL-11	Platform shall ensure that the pseudonymized data can be amended or withdrawn after its sharing, if the data subject (patient) requests the modification.		Legal