Skip to content

VUB brings expertise in AI and ethics to DIGILUNG project

The Vrije Universiteit Brussel (VUB), a leading Urban Engaged University located in the heart of Europe, is contributing cutting-edge expertise to the DIGILUNG project. VUB is committed to solving real-world challenges through interdisciplinary research and innovation. Learn more at www.vub.be.

 

Bridging AI, Ethics, and Clinical Relevance

Although VUB joined the DIGILUNG consortium during the project’s preparation phase, its contributions have proven vital. The university was invited to participate due to its dual expertise: the FARI institute’s leadership in ethics and legal aspects of AI, and VUB’s strong technical background in image and signal processing – both essential to DIGILUNG’s objectives.

 

VUB

Building on Past Achievements

VUB’s involvement builds on a strong foundation of previous research in medical imaging and AI. The university has developed advanced methods for AI-driven medical image registration and lesion classification, including work on COVID-19 diagnostics. These capabilities are now being adapted for more robust and clinically relevant tools for detecting and monitoring lung diseases within the DIGILUNG project.

Moreover, VUB brings valuable experience in ethical AI and medical regulation, currently applied to guide the ethical-legal framework that supports DIGILUNG’s compliance with medical standards and prepares it for clinical implementation.

A unique aspect of VUB’s expertise comes from previous research into gait analysis using inertial sensors. These AI-based methods, originally designed to detect early signs of motor impairments in elderly and neurological populations, are now being repurposed for the analysis of respiratory signals from wearable devices like the LungShirt. This interdisciplinary approach demonstrates how foundational work in one field can successfully drive innovation in another.

Key Contributions in 2024

In 2024, VUB made significant progress on both ethical-legal and technical fronts:

  • Ethical-Legal Framework: VUB led the documentation and alignment of the DIGILUNG project with relevant legal and ethical standards. A comprehensive report was developed in collaboration with consortium partners to ensure the project’s integrity and regulatory readiness.
  • Medical Imaging: VUB introduced an advanced AI methods for analyzing lung scans of COPD patients. After identifying inconsistencies in existing methods, the team implemented a new AI-based approach that improved accuracy and enabled regional lung analysis – offering better insight into COPD subtypes. Tools were also developed to align longitudinal scans from the same patient, aiding in tracking disease progression.
  • Breathing Data Analysis: Using knowledge from gait research, VUB developed and validated AI-based solution for denoising respiratory signals collected by wearable sensors. These tools significantly improved the clarity of breathing data, making it easier to extract clinically meaningful metrics like breath rate and volume. The methods have been shared with ATRIDE for further development.

Looking Ahead to 2025

VUB’s work in 2025 will focus on two key areas:

  • Ethical-Legal Progress: The team will conduct an impact assessment and validation among all partners. This will initiate iterative discussions between legal and technical representatives of the partners on requirements and how they can be implemented in practice. The findings will be used to create the diagnostics of data needed, the required changes, and will feed into social acceptance and socio-economic impact assessment of the products.
  • Technical Development: VUB will further refine its AI-based methods for detecting anomalies in lung medical images. In parallel, the team will finalize the RIP belt signal denoising tools, this time using data from both healthy individuals and patients.

 

The project is funded by the European Union. The expressed views and opinions belong solely to the author(s) and may not necessarily reflect the views of the European Union or EISMEA. Neither the European Union nor the funding institution can be held responsible for the expressed views.