Periodic Reporting for period 2 - LUCIA (Understanding Lung Cancer related risk factors and their Impact)
Berichtszeitraum: 2024-07-01 bis 2025-12-31
Lung cancer remains the leading cause of cancer-related death in Europe. While smoking is a major risk factor, many patients develop lung cancer without clear or well-quantified exposure history. Current screening approaches rely primarily on imaging (low-dose CT), which can be costly, resource-intensive, and associated with overdiagnosis.
LUCIA (Lung Cancer Risk Identification and Assessment) aims to improve the understanding and quantification of lung cancer risk by integrating multiple complementary domains: personal and lifestyle factors, environmental exposures, and biological responses at molecular and cellular levels. The project develops and validates a multimodal “LUCIA toolbox” combining non-invasive sensing technologies, artificial intelligence (AI)-based risk modelling, large-scale clinical datasets, and socio-ethical guidance.
LUCIA does not implement screening programmes directly. Instead, it provides scientifically validated methods, interoperable datasets, and decision-support tools that may inform future evidence-based prevention, screening, and early detection strategies.
LUCIA (Lung Cancer Risk Identification and Assessment) aims to improve the understanding and quantification of lung cancer risk by integrating multiple complementary domains: personal and lifestyle factors, environmental exposures, and biological responses at molecular and cellular levels. The project develops and validates a multimodal “LUCIA toolbox” combining non-invasive sensing technologies, artificial intelligence (AI)-based risk modelling, large-scale clinical datasets, and socio-ethical guidance.
LUCIA does not implement screening programmes directly. Instead, it provides scientifically validated methods, interoperable datasets, and decision-support tools that may inform future evidence-based prevention, screening, and early detection strategies.
During the second reporting period, LUCIA transitioned from initial platform development to large-scale clinical data generation and operational validation.
Key achievements include:
1. Continued recruitment across multiple European clinical sites, expanding both retrospective and prospective datasets.
2. Deployment and use of non-invasive sensing technologies, including a breath analyser (BAN), a skin-based sensing patch (WBSP), and spectrometry-based devices (SPOC), within ongoing clinical studies.
3. Consolidation of a secure Health Data Platform aligned with the OMOP Common Data Model, enabling harmonised multi-site data integration.
4. Operationalisation of a Virtual Research Environment supporting secure AI-based modelling across clinical partners.
5. Development and validation of exploratory multimodal AI models integrating clinical, imaging, molecular, and sensor-derived data.
6. Ongoing stakeholder engagement through Social Labs and structured ethical oversight to support trustworthy and socially responsible AI development.
These activities generated the first cross-site real-world validation evidence for the LUCIA toolbox under operational clinical conditions.
Key achievements include:
1. Continued recruitment across multiple European clinical sites, expanding both retrospective and prospective datasets.
2. Deployment and use of non-invasive sensing technologies, including a breath analyser (BAN), a skin-based sensing patch (WBSP), and spectrometry-based devices (SPOC), within ongoing clinical studies.
3. Consolidation of a secure Health Data Platform aligned with the OMOP Common Data Model, enabling harmonised multi-site data integration.
4. Operationalisation of a Virtual Research Environment supporting secure AI-based modelling across clinical partners.
5. Development and validation of exploratory multimodal AI models integrating clinical, imaging, molecular, and sensor-derived data.
6. Ongoing stakeholder engagement through Social Labs and structured ethical oversight to support trustworthy and socially responsible AI development.
These activities generated the first cross-site real-world validation evidence for the LUCIA toolbox under operational clinical conditions.
LUCIA advances beyond single-modality approaches by:
1. Combining non-invasive breath and skin sensing with clinical and molecular risk information.
2. Developing multimodal AI workflows that can integrate imaging, omics, and non-invasive sensor outputs within a common research infrastructure.
3. Enabling cross-site analysis through a secure data ecosystem supported by data governance and GDPR-relevant controls.
4. Using synthetic data resources (where applicable) to support technical development while reducing privacy risks.
5. Embedding ethical, legal and societal aspects (ELSA) and AI impact assessment within the project lifecycle.
Overall, LUCIA establishes a structured framework for integrated lung cancer risk assessment that complements current imaging-led screening approaches.
1. Combining non-invasive breath and skin sensing with clinical and molecular risk information.
2. Developing multimodal AI workflows that can integrate imaging, omics, and non-invasive sensor outputs within a common research infrastructure.
3. Enabling cross-site analysis through a secure data ecosystem supported by data governance and GDPR-relevant controls.
4. Using synthetic data resources (where applicable) to support technical development while reducing privacy risks.
5. Embedding ethical, legal and societal aspects (ELSA) and AI impact assessment within the project lifecycle.
Overall, LUCIA establishes a structured framework for integrated lung cancer risk assessment that complements current imaging-led screening approaches.