Sahar Almahfouz Nasser

Sahar Almahfouz Nasser

Postdoctoral Researcher, Emory University. PhD in Electrical Engineering from IIT-Bombay

About Me

I am a postdoctoral researcher in the Department of Biomedical Engineering at Emory University, working with Dr. Anant Madabhushi on computational pathology and AI for precision oncology. My expertise spans machine learning, computer vision, and medical image analysis, with publications in venues such as WACV, MICCAI, ISBI, ASCO, JCO, and Ultrasonics.

I completed my PhD at the Indian Institute of Technology Bombay, guided by Prof. Amit Sethi, where I worked on cross-domain image adaptation and matching. My doctoral training combined digital signal processing and machine learning with precision medicine, conducted in close collaboration with clinicians at Tata Memorial Hospital.

At Emory, I have advanced robust mitosis detection models, explainable AI tools for HER2 scoring that outperform FDA-approved assays, and prognostic models for ER+/HER2− breast cancer. My recent work emphasizes agentic AI systems for oncology, including multi-agent frameworks for tumor microenvironment phenotyping, federated learning pipelines for secure collaboration, and hallucination-free image normalization across modalities.

My long-term vision is to develop explainable, generalizable, and equitable AI-driven biomarkers that personalize treatment decisions, reduce overtreatment, and make precision cancer care more accessible worldwide.


I maintain a list of my publications under the Research tab and my full Resume is attached the CV Tab.

News

June 2025 — New Paper Published: Our paper on HAI-score, an objective AI method for HER2 H-score estimation from IHC breast cancer samples, is now published.

June 2025 — Clinical Trial Biomarker: Published a study on a computational pathology–informed immune biomarker for trastuzumab benefit in HER2+ breast cancer, validated in the NSABP B-41 trial.

March 2025 — Perspective Article: Opportunities for Artificial Intelligence in Oncology: From the Lens of Clinicians and Patients was published.

2024 — Conference Presentation: Presented our work on radiomic features for MRI keypoint detection at BIOSTEC 2024.

2024 — Accepted Paper: Our paper on Reverse Knowledge Distillation for retinal image matching was accepted at WACV 2024.

August 2023 — Preprint Released: Transforming Breast Cancer Diagnosis: Towards Real-Time Ultrasound to Mammogram Conversion is now on arXiv.

July 2023 — Research Highlight: Our Reverse Knowledge Distillation approach became state-of-the-art for retinal image registration.

March 2023 — Best Paper Award: Our UNet-Based Adversarial Domain Homogenizer for mitosis detection won Best Paper at BIOSTEC 2023.