B2SLab pursues research across five strategic goal areas. Below is an overview of our main current and recent projects, organised by research goal. Note that industrial doctorate projects appear both in their thematic section and in Technology Transfer.
Foundational Models in Health
Building large-scale AI models trained on real-world clinical data is a core strategic goal of B2SLab. We believe that foundation models — pre-trained on broad, heterogeneous health data and fine-tuned for specific tasks — represent the most promising path toward AI that genuinely improves clinical care. Our work spans electronic health records and emergency medicine data.
Deep learning for Electronic Health Records · Ministerio de Ciencia (AEI)
Transformer-based models trained on data from over 200,000 patients (SIDIAP database) to predict disease trajectories in Type 2 Diabetes and COPD. This line of work produced the DARE model and is now evolving toward a full diabetes foundation model.
Read more →
Multimodal foundation model for Emergency Department data · Blanca Aleajos · Partner: Hospital Sant Joan de Déu · Industrial Doctorate
Building a multimodal foundation model trained on real-world Emergency Department clinical data — a reusable AI infrastructure for clinical decision support and operational planning at scale.
Read more →
Rare Diseases
Understanding and characterising rare diseases is a long-standing commitment of B2SLab. Our work combines patient data platforms, computational phenotyping, and in silico modelling to improve diagnosis and accelerate research for conditions that affect small — but deeply underserved — patient populations.
Patient data platform for rare diseases · EU Horizon 2020 #780262 · Project concluded, platform active
A European platform connecting 5,000+ rare disease patients and caregivers across 50+ countries. B2SLab designed and built the entire data science backend — including HPO-based phenotype standardisation and the People Like Me similarity engine — and continues to maintain the platform’s data science infrastructure.
Read more →
Paediatric neurometabolic disorders · Ministerio de Ciencia (AEI)
Age-adjusted reference intervals for CSF amino acids and serum copper in large paediatric cohorts, in collaboration with Dr. Rafael Artuch at Hospital Sant Joan de Déu.
Read more →
Computational evaluation of experimental disease models · Aitor Moruno · Partner: Almirall SA · Industrial Doctorate
Computational methods — including the singIST tool — to quantify how faithfully preclinical animal models reproduce human skin disease biology at single-cell resolution, enabling principled model selection before drug testing begins.
Read more →
Technology Transfer
B2SLab is deeply committed to translating research into real-world impact. This includes coordinating a major regional health technology transfer network and supervising industrial doctorate students embedded in partner companies and hospitals, where academic research is developed and validated in operational settings.
Catalan HealthTech Transfer Network · Active since 2018 · Ref. 2025 XARDI 00008 · ERDF / Generalitat de Catalunya / AGAUR
A network of 122 research groups from 31 institutions, co-funded by the European Regional Development Fund, dedicated to accelerating the transfer of medical technologies from research to market across Catalonia. The network is coordinated by Prof. Perera Lluna and organises the annual HealthTech2030 summit and Open Innovation Challenge.
Read more →
Multimodal foundation model for Emergency Department data · Blanca Aleajos · Partner: Hospital Sant Joan de Déu · Industrial Doctorate
Building a multimodal foundation model trained on real-world Emergency Department clinical data — a reusable AI infrastructure for clinical decision support and operational planning at scale.
Read more →
Computational evaluation of experimental disease models · Aitor Moruno · Partner: Almirall SA · Industrial Doctorate
Computational methods — including the singIST tool — to quantify how faithfully preclinical animal models reproduce human skin disease biology at single-cell resolution, enabling principled model selection before drug testing begins.
Read more →
Climate Change & Sensors
Climate change poses direct and measurable risks to human health. B2SLab develops AI platforms that link environmental and meteorological data to clinical outcomes, deploys low-cost sensor networks for environmental monitoring in vulnerable ecosystems, and maintains a broad programme of chemosensor and air quality research.
AI for extreme climate events and chronic disease risk · Active · Hospital pilot: Mútua Terrassa
An AI platform integrating meteorological, environmental, and clinical data to anticipate the impact of extreme heat episodes on patients with chronic diseases. Currently in hospital pilot phase at Mútua Terrassa University Hospital.
Read more →
Greenhouse gas regulation in headwater catchments · Water4All / AEI · Active from 2025
An interdisciplinary European project (Spain, Finland, Sweden, Italy, Czech Republic) investigating how climate change, nutrient deposition, and land use affect headwater catchments’ capacity to regulate CO₂, CH₄, and N₂O. B2SLab designs, builds, and deploys the low-cost multi-sensor monitoring nodes — now operational at the Furiosos catchment in Catalunya.
Read more →
Electronic nose, CO₂ monitoring, and personal exposure · Jordi Fonollosa
A programme of sensor development and deployment spanning electronic noses for food quality control, wearable air quality monitors for asthma patients, CO₂ sensor placement guidelines for schools, and the award-winning Nepal indoor air quality project (UPC Social Commitment Prize 2025).
Read more →
For the full list of publications associated with our projects, visit the Publications page.