Periodic Reporting for period 2 - OPTIMA (Optimal treatment for patients with solid tumours in Europe through Artificial intelligence)
Reporting period: 2022-10-01 to 2023-09-30
OPTIMA is a multi-stakeholder-led consortium with the vision that each and every patient should have access to the most up-to-date individualised treatments and to innovative therapies. In that perspective, we will design, develop and deliver the first interoperable and GDPR compliant European real-world oncology data and evidence generation platform, along with decision support toolsets for prostate, lung and breast cancer. By designing this platform from the onset based on the needs of the clinicians and patients in an inclusive and sustainable way, OPTIMA aims to strengthen shared decision-making based on innovative data and AI-driven technology and tools.
To achieve our vision, we have united the key partners from other data-driven European wide oncology initiatives (PIONEER, EHDEN, HARMONY) in our project along with Europe's leading clinicians in the fields of breast, prostate and lung cancer that all have leading roles in medical societies such as ERS, ESMO, and EAU. At the core of the OPTIMA project are the major guideline offices of the three indications, whilst the project is further strengthened by patient advocacy groups, well-known academic partners and 6 SMEs as partners covering the needed expertise to implement OPTIMA successfully. The platform will build on existing open source solutions such as the OMOP-CDM, tranSMART and OHDSI ATLAS whilst developing novel AI models based on federated learning to facilitate knowledge discovery. At the start of the project, we will already have access to datasets covering over 200M people in Europe, along with data from high quality patient cohorts for AI model building. Moreover, the platform will be hosted in-kind on existing infrastructure provided by Helmholtz Institute in Germany. Finally, OPTIMA envisions sustainability of its platform through a value-based approach with various business models for commercial users.
To achieve our vision, we have united the key partners from other data-driven European wide oncology initiatives (PIONEER, EHDEN, HARMONY) in our project along with Europe's leading clinicians in the fields of breast, prostate and lung cancer that all have leading roles in medical societies such as ERS, ESMO, and EAU. At the core of the OPTIMA project are the major guideline offices of the three indications, whilst the project is further strengthened by patient advocacy groups, well-known academic partners and 6 SMEs as partners covering the needed expertise to implement OPTIMA successfully. The platform will build on existing open source solutions such as the OMOP-CDM, tranSMART and OHDSI ATLAS whilst developing novel AI models based on federated learning to facilitate knowledge discovery. At the start of the project, we will already have access to datasets covering over 200M people in Europe, along with data from high quality patient cohorts for AI model building. Moreover, the platform will be hosted in-kind on existing infrastructure provided by Helmholtz Institute in Germany. Finally, OPTIMA envisions sustainability of its platform through a value-based approach with various business models for commercial users.
WP1: Project Management – WP1 has regular meetings to actively monitor the progress and identify and mitigate any potential risks for the project. WP1 has also delivered the first two periodic reports and organised a virtual kick-off meeting and two face-to-face consortium meetings in June 2022 and May 2023.
WP2: Informed consent, general requirement analysis, governance, and regulation – The WP has assessed the legal and ethical framework of the project, as well as the data protection impact, which serve as important prerequisites for any further work within the OPTIMA project. The Art. 26/28 agreements have been drafted and signed by all beneficiaries and WP2 is now drafting the external data sharing agreements.
WP3: Guideline-based decision support tool – The WP has been identifying and prioritizing critical research questions through Delphi surveys with key opinion leaders and stakeholders, and selecting typical case presentations for prostate, breast and lung cancer, which will serve as the basis for the development of the guideline-based decision support tool.
WP4: Platform technical and infrastructural requirement analysis – The WP has so far worked on the initial data management plan, assessment of the requirements and specifications of the platform, and on the development environment. The first release candidate of the platform has been developed and improved after initial tests and is currently being installed and tested at EMC and ICO.
WP5: Platform implementation and evaluation – The WP has been working on identifying the user personas and has started working on implementing the platform in the clinical sites. WP5 have supported with the implementation of the platform and tested it for stability and function. The WP will continuously search for additional sites where version 2 and 3 can be tested.
WP6: (Non)interventional study data and real-world data gathering, preparation and integration – The WP has so far been mapping potential datasets, designed the architecture of federated nodes and decided on the common data model approach used. This WP has developed a fingerprinting questionnaire to assess the data governance, helped in the design of the architecture of the federated notes, assessed the security of data AI-models, organized trainings on semi-automated tools for OMOP-mapping, and decided on a strategy to convert datasets from the consortium partners into OMOP.
WP7: Artificial intelligence knowledge base implementation – The WP has been investigating the use of synthetic data, assessing available AI/ML models, explored the options for federated learning and commenced the work around NLP. WP7 has developed an overview of the state of the art for AI models in the three cancer types investigated in OPTIMA and has started the development of a prototype AI model relevant to answer research questions that can benefit from automatic analysis of imaging data. WP7 has also provided input to assess the impact of AI both for answering the research questions as well as business cases for the sustainability beyond the funding phase. In terms of infrastructure, collection of AI models and a working CI/CD environment are now set up.
WP8: Dissemination and communication – focuses on dissemination and communication of the project’s aims, approach and results, strengthening stakeholder engagement. The WP has established the house-style, developed needed templates and managed OPTIMA social media accounts. WP8 have created a network with relevant stakeholders
WP9: Exploitation and sustainability of the results – aims to build a sound and solid business strategy to exploit and commercialize the platform and the guideline decision support tool beyond the grant period. The WP has conducted interviews to assess scientific requirements, benchmarked potential business models and starting defining models for commercialization. A sustainability workshop was held in September 2022. In 2023, WP9 has collaborated with the other WPs and conducted workshops to further define a sustainable model for OPTIMA.
WP2: Informed consent, general requirement analysis, governance, and regulation – The WP has assessed the legal and ethical framework of the project, as well as the data protection impact, which serve as important prerequisites for any further work within the OPTIMA project. The Art. 26/28 agreements have been drafted and signed by all beneficiaries and WP2 is now drafting the external data sharing agreements.
WP3: Guideline-based decision support tool – The WP has been identifying and prioritizing critical research questions through Delphi surveys with key opinion leaders and stakeholders, and selecting typical case presentations for prostate, breast and lung cancer, which will serve as the basis for the development of the guideline-based decision support tool.
WP4: Platform technical and infrastructural requirement analysis – The WP has so far worked on the initial data management plan, assessment of the requirements and specifications of the platform, and on the development environment. The first release candidate of the platform has been developed and improved after initial tests and is currently being installed and tested at EMC and ICO.
WP5: Platform implementation and evaluation – The WP has been working on identifying the user personas and has started working on implementing the platform in the clinical sites. WP5 have supported with the implementation of the platform and tested it for stability and function. The WP will continuously search for additional sites where version 2 and 3 can be tested.
WP6: (Non)interventional study data and real-world data gathering, preparation and integration – The WP has so far been mapping potential datasets, designed the architecture of federated nodes and decided on the common data model approach used. This WP has developed a fingerprinting questionnaire to assess the data governance, helped in the design of the architecture of the federated notes, assessed the security of data AI-models, organized trainings on semi-automated tools for OMOP-mapping, and decided on a strategy to convert datasets from the consortium partners into OMOP.
WP7: Artificial intelligence knowledge base implementation – The WP has been investigating the use of synthetic data, assessing available AI/ML models, explored the options for federated learning and commenced the work around NLP. WP7 has developed an overview of the state of the art for AI models in the three cancer types investigated in OPTIMA and has started the development of a prototype AI model relevant to answer research questions that can benefit from automatic analysis of imaging data. WP7 has also provided input to assess the impact of AI both for answering the research questions as well as business cases for the sustainability beyond the funding phase. In terms of infrastructure, collection of AI models and a working CI/CD environment are now set up.
WP8: Dissemination and communication – focuses on dissemination and communication of the project’s aims, approach and results, strengthening stakeholder engagement. The WP has established the house-style, developed needed templates and managed OPTIMA social media accounts. WP8 have created a network with relevant stakeholders
WP9: Exploitation and sustainability of the results – aims to build a sound and solid business strategy to exploit and commercialize the platform and the guideline decision support tool beyond the grant period. The WP has conducted interviews to assess scientific requirements, benchmarked potential business models and starting defining models for commercialization. A sustainability workshop was held in September 2022. In 2023, WP9 has collaborated with the other WPs and conducted workshops to further define a sustainable model for OPTIMA.
Our vision is that every patient should have access to the most up-to-date individualised treatments and to innovative therapies. By strengthening shared decision-making based on dynamic computer-interpretable guidelines (CIGs), innovative broad data access and AI-driven technology and tools, we envision revolutionizing oncology care in Europe. This vision is realized by building the first interoperable and GDPR-compliant real-world-oncology data and evidence generation platform based on the needs of the clinicians and patients, and a comprehensive decision support toolsets based on national and international guidelines with approved regular updates of guideline recommendations underpinned by evidence from advanced statistical analysis and AI. Our ambition is to scale up OPTIMA to a pan-European level collaborating with national and international medical societies’ guidelines addressing a paradigm shift in clinical research and treatment towards a more personalised medicine approach ensuring that each patient receives the best individualised treatment. We envision that the integration of individual patient data will enable personalised decisions in near real time while continuously enhancing the quality and capabilities of the platform provided.