Primary Research Areas
Digital Pathology
Digital pathology represents a paradigm shift in how we analyze tissue samples. My research focuses on:
- Whole-Slide Image Analysis: Algorithms and architectures for gigapixel-resolution image interpretation
- Tissue Classification: Automated identification and segmentation of tissue types and pathological regions
- Diagnostic Support: Creating AI systems that enhance pathologist productivity and reduce errors
Computational Pathology
Bridging AI and clinical pathology:
- Cancer Detection & Grading: Automated systems for cancer identification and prognostic assessment
- Biomarker Discovery: Computational methods for identifying clinically relevant markers
- Clinical Integration: Designing AI tools that integrate seamlessly into clinical workflows
Medical Image Analysis
Developing robust deep learning approaches:
- Limited Annotation Learning: Techniques that work with sparse annotations in large image datasets
- Multi-scale Analysis: Approaches that leverage information at different image resolutions
- Domain Adaptation: Methods to transfer models across different data sources and institutions
Featured Projects
Automated Prostate Cancer Grading
Developing deep learning methods for automated Gleason score prediction. This project addresses standardization challenges in cancer grading and improves interobserver reproducibility.
Deep Learning for Whole-Slide Image Analysis
Building attention-based architectures for analyzing complete histopathological slides, combining global context with local detail analysis.
Multi-Instance Learning for Pathology
Exploring weakly supervised learning frameworks that can learn from limited labels in large-scale medical image datasets.
Interested in Our Work?
Collaboration opportunities: Academic partnerships, industry collaborations, and clinical validation studies
Student recruitment: PhD positions and postdoctoral fellowships available
- Clinical Trials: Prospective studies to validate clinical utility
Future Directions
Looking forward, we aim to:
- Develop multimodal AI systems that combine histopathology with genomic and clinical data
- Create explainable AI models that enhance pathologist decision-making
- Expand applications to other cancer types and organs
- Establish best practices for clinical deployment of pathology AI