Mohamed Harmanani mharmanani

👋 Hello, I'm Mohamed

🔬 About Me

I am a Doctoral Researcher at the Vector Institute and a PhD student at Queen's University in the Medical Informatics Laboratory. My research focuses on designing, training, and deploying trustworthy deep learning models for disease diagnosis and computer-assisted interventions.

My research interests include:

🧠 Deep Learning
🩻 Medical Imaging
🖼️ Computer Vision
💬 Natural Language Processing

Most of my repositories reflect ongoing or completed projects, tools, or experiments related to my PhD work.

📚 Research Highlights

🔬 Current Projects

🧠 Image-Guided Medical Reasoning – Using large vision-language and reasoning models for interpretable diagnostics in medical imaging.
🎯Attribute Localization and Visual Grounding – Identifying diagnostically relevant regions in medical images using attention and visual explanations.
📊 Uncertainty Quantification – Modeling and estimating predictive uncertainty in AI-assisted diagnostics to improve reliability, calibration, and clinical trustworthiness.
🔄 Multi-Modal Knowledge Transfer – Learning from diverse data and imaging modalities (e.g., histology, MRI) and transferring knowledge to low-data clinical settings.