Periodic Reporting for period 1 - VASCUL-AID (DEVELOPING TRUSTWORTHY ARTIFICIAL INTELLIGENCE (AI)-DRIVEN TOOLS TO PREDICT VASCULAR DISEASE RISK AND PROGRESSION)
Reporting period: 2023-05-01 to 2024-10-31
The aim of VASCUL-AID is to predict the risk of cardiovascular events and progression of the vascular diseases Abdominal Aortic Aneurysm (AAA) and Peripheral Arterial Disease (PAD) to influence the course of disease improving the patient’s quality of life and care and assisting clinicians to make better-informed decisions involving the patient. VASCUL-AID will allow us for the first time to identify patients who are at high risk for AAA growth or PAD progression and cardiovascular events.
To this end, we will deliver a clinically relevant and cost-effective trustworthy AI-driven platform (VASCUL-AID) that integrates multi-source parameters including imaging, proteomic and genomic data as well as life-style patient data from wearables to enable personalised vascular disease management.
To maximise the personalised prevention strategies, VASCUL-AID leverages visualisation tools to improve clinician patient communication and empower the patient. The VASCUL-AID platform consists of AI risk-prediction tools, a patient communication app an a clinical dashboard to support clinical decision-making. A particular emphasis is placed on ethics, to ensure beneficial implementation of AI prediction tools.
In this project, we aim to:
(1) build an EU-wide data infrastructure
(2) develop an AI-based progression prediction tools for AAA and PAD
(3) develop criteria according to the COMET initiative to assess the effectiveness of VASCUL-AID, and
(4) clinically test and show proof-of-concept for the VASCUL-AID platform.
Once validated, this platform can be extended to other cardiovascular diseases (CVDs) as well. VASCUL-AID brings together 14 leading organisations (and 2 affiliated entities) consisting of clinical academic centres, institutes, universities and SMEs as well as large industry, patients organisations and policy makers that cover the full value chain to enable integration of the platform into clinical practice.
To this end, we will deliver a clinically relevant and cost-effective trustworthy AI-driven platform (VASCUL-AID) that integrates multi-source parameters including imaging, proteomic and genomic data as well as life-style patient data from wearables to enable personalised vascular disease management.
To maximise the personalised prevention strategies, VASCUL-AID leverages visualisation tools to improve clinician patient communication and empower the patient. The VASCUL-AID platform consists of AI risk-prediction tools, a patient communication app an a clinical dashboard to support clinical decision-making. A particular emphasis is placed on ethics, to ensure beneficial implementation of AI prediction tools.
In this project, we aim to:
(1) build an EU-wide data infrastructure
(2) develop an AI-based progression prediction tools for AAA and PAD
(3) develop criteria according to the COMET initiative to assess the effectiveness of VASCUL-AID, and
(4) clinically test and show proof-of-concept for the VASCUL-AID platform.
Once validated, this platform can be extended to other cardiovascular diseases (CVDs) as well. VASCUL-AID brings together 14 leading organisations (and 2 affiliated entities) consisting of clinical academic centres, institutes, universities and SMEs as well as large industry, patients organisations and policy makers that cover the full value chain to enable integration of the platform into clinical practice.
The VASCUL-AID project has made significant advances across its work packages, ensuring adequate management, innovative research, and impactful dissemination.
In WP1: Project Management and Governance, regular consortium and work package-specific meetings were conducted, including a kick-off in Amsterdam and an annual in-person meeting in Porto. A comprehensive project management handbook and risk mitigation protocols were developed to guide the initiative effectively.
In WP2: Data Collection and Standardization, a standardized repository for AAA and PAD data was established, integrating retrospective datasets and biobank samples. Ethical approvals were secured for retrospective studies, while prospective studies are nearing approval. Protocols and SOPs for patient data collection were finalized, alongside collaborations with biobanks. Standard reporting outcome sets and data dictionaries for AAA and PAD were developed, with mid-term recruitment reports submitted for over 10,000 patient records.
WP3 focused on AI Tools for AAA Progression Prediction, including the development of a gene-protein ontology prediction model and the integration of proteomics and lipidomics results. Automated segmentation models for aortic structures and geometric deep learning methodologies for hemodynamic analyses were advanced.
Similarly, in WP4: AI Tools for PAD Progression Prediction, segmentation algorithms for peripheral artery imaging were completed, and predictive models integrating longitudinal clinical and -omics data are underway.
The first prototypes of patient and clinician apps were delivered in WP5, incorporating wearable integration and visualization tools for communication and decision-making. Prospective clinical validation study designs were finalized in WP6, while WP7 ensured alignment with EU AI Act requirements through the HELSI AI framework and ethical workshops. In WP8, dissemination activities included launching the VASCUL-AID website, presenting findings at conferences, and fostering dialogue with stakeholders.
In WP1: Project Management and Governance, regular consortium and work package-specific meetings were conducted, including a kick-off in Amsterdam and an annual in-person meeting in Porto. A comprehensive project management handbook and risk mitigation protocols were developed to guide the initiative effectively.
In WP2: Data Collection and Standardization, a standardized repository for AAA and PAD data was established, integrating retrospective datasets and biobank samples. Ethical approvals were secured for retrospective studies, while prospective studies are nearing approval. Protocols and SOPs for patient data collection were finalized, alongside collaborations with biobanks. Standard reporting outcome sets and data dictionaries for AAA and PAD were developed, with mid-term recruitment reports submitted for over 10,000 patient records.
WP3 focused on AI Tools for AAA Progression Prediction, including the development of a gene-protein ontology prediction model and the integration of proteomics and lipidomics results. Automated segmentation models for aortic structures and geometric deep learning methodologies for hemodynamic analyses were advanced.
Similarly, in WP4: AI Tools for PAD Progression Prediction, segmentation algorithms for peripheral artery imaging were completed, and predictive models integrating longitudinal clinical and -omics data are underway.
The first prototypes of patient and clinician apps were delivered in WP5, incorporating wearable integration and visualization tools for communication and decision-making. Prospective clinical validation study designs were finalized in WP6, while WP7 ensured alignment with EU AI Act requirements through the HELSI AI framework and ethical workshops. In WP8, dissemination activities included launching the VASCUL-AID website, presenting findings at conferences, and fostering dialogue with stakeholders.
The development and implementation of the VASCUL-AID platform have led to advancements in vascular healthcare, co-designed with patients, clinicians, AI experts, policymakers, and other stakeholders to align with the EU AI Act. Key innovations include standardized data dictionaries and outcome sets for AAA and PAD, which ensure consistent data collection and interoperability. AI-driven risk prediction tools are developed to provide accurate, personalized assessments, enabling clinicians to optimize care strategies. A European-wide standardized data infrastructure facilitates secure and compliant data sharing, while medical-grade applications enhance patient-clinician interactions by delivering actionable insights in an accessible format. A trustworthy AI framework ensures transparency and adherence to ethical standards, with advanced visualization tools that present complex health data in an intuitive manner.
Patients and clinicians are empowered through tools that visualize health metrics and risks, fostering informed decision-making and proactive healthcare engagement. Research capabilities are significantly advanced, with standardized, traceable data collection paving the way for the platform to address other cardiovascular conditions. Moreover, by incorporating robust ethical guidelines and regulatory compliance mechanisms, VASCUL-AID sets a benchmark for AI in healthcare and has the potential to shape future EU policies.
To ensure further uptake and success, ongoing research, demonstration, and clinical validation are essential. Scaling commercialization efforts, securing market access, and expanding internationally will be critical. Additionally, supportive regulatory frameworks and harmonized standards across the EU will play a vital role in sustaining the platform's transformative impact on vascular healthcare.
Patients and clinicians are empowered through tools that visualize health metrics and risks, fostering informed decision-making and proactive healthcare engagement. Research capabilities are significantly advanced, with standardized, traceable data collection paving the way for the platform to address other cardiovascular conditions. Moreover, by incorporating robust ethical guidelines and regulatory compliance mechanisms, VASCUL-AID sets a benchmark for AI in healthcare and has the potential to shape future EU policies.
To ensure further uptake and success, ongoing research, demonstration, and clinical validation are essential. Scaling commercialization efforts, securing market access, and expanding internationally will be critical. Additionally, supportive regulatory frameworks and harmonized standards across the EU will play a vital role in sustaining the platform's transformative impact on vascular healthcare.