Artificial Intelligence Scientist at Unilabs &
Invited Assistant Professor at FEUP.
Bridging the gap between Academic Research and Clinical Application.
Focused on interpretable Machine Learning and Computer-Aided Diagnosis.
Novel Multi-label Aggregation Strategy for Multiple-Instance Learning Tasks in Digital Pathology
Developed a novel semi-supervised approach to leverage weak and full-annotations for an efficient MIL approach with tile sampling. Released +5000 medical exams dataset with ~7TB of data.
Introduced and aggregated knowledge developed across a large digital pathology project, to create guidelines on how to close the gap between engineers and pathologists for efficient and accurate annotation.
Coverage of our AI prototype for colorectal cancer diagnosis.
TV news report featuring our work on AI-assisted diagnostics.
International coverage of our research breakthrough.
An open discussion on the national TV regarding what is AI and how it is impacting different aspects of healthcare.
An analysis of the current state of AI in healthcare, distinguishing between market hype and genuine clinical value.
A discussion on the mitigation of face recognition demographic bias.
Unilabs • 2024 - Present
FEUP • 2022 - Present
INESC TEC • 2020 - 2024
Feedzai • 2020 - 2020
Aalto University • 2019 - 2020
Artificial Intelligence Solution to Diagnose Colorectal Cancer and Dysplasia from Whole-Slide Images.
Novel Explainable Artificial Intelligence approaches applied to Biometrics and Face Recognition tasks. Most are generalizable across domains.
Developing Medical AI at Unilabs across a diverse spectrum of data types, domains and requirements.
PyTorch, Scikit-learn, Computer Vision, Deep Learning.
Python, C++, SQL
Git, Docker, LaTeX, Digital Pathology (WSI), Biometrics.
Medical Imaging, Computer-aided Diagnosis, Explainable AI (XAI).