Blanca Alaejos Pardo
foundation models, emergency medicine, emergency admissions, EHR, multimodal learning, clinical AI, Sant Joan de Deu, UPC, industrial doctorate
Bio
Blanca Alaejos Pardo is an industrial doctorate researcher at B2SLab / IRIS-UPC. Her PhD work is carried out in collaboration with Hospital Sant Joan de Deu, where she works on foundation models for Emergency Department admissions.
Her research focuses on building reusable AI models from real-world emergency care data. The goal is to learn useful patient representations from the heterogeneous information generated during an Emergency Department visit, including structured clinical variables, temporal event sequences, categorical codes, and clinical text.
Research
Blanca’s doctoral project, Foundation Models for Emergency Medicine, explores how multimodal foundation models can support clinical decision-making and operational planning in emergency admissions.
The project studies how to pre-train a general-purpose model on Emergency Department data and adapt it to downstream tasks such as:
- patient disposition prediction,
- deterioration or risk stratification,
- length-of-stay estimation,
- resource demand forecasting,
- reusable clinical representations for future AI applications.
This work is part of B2SLab’s broader research programme on foundation models for health, extending the group’s experience with transformer-based models for longitudinal Electronic Health Records into paediatric emergency medicine.
Project
| Doctorate type | Industrial Doctorate |
| Research line | Multimodal foundation models for Emergency Department data |
| Clinical partner | Hospital Sant Joan de Deu, Barcelona |
| University | Universitat Politecnica de Catalunya (UPC) |
| Research group | B2SLab / IRIS-UPC |
| Supervisor | Alexandre Perera Lluna |
Publications
Blanca’s publication list will be added here as soon as her publications are indexed in FUTUR or Google Scholar.