Portrait of Noel Codella.

Principal Researcher

Microsoft

Noel C. F. Codella

Noel C. F. Codella is a Principal Researcher at Microsoft whose work sits at the intersection of foundation models, computer vision, and healthcare AI. He has made influential contributions spanning multimodal vision-language learning, transformer architectures, representation learning, fairness and evaluation, and clinically grounded medical AI, with publications in venues including CVPR, ICCV, ECCV, ICLR, MICCAI, AAAI, Nature Medicine, and Lancet Oncology.

His technical foundation includes graduate work in cardiac MRI spanning imaging physics, data acquisition, reconstruction, and downstream clinical analysis - a rare combination that informs his approach to building robust and trustworthy medical AI systems. At Microsoft, he has helped drive frontier work in both general computer vision and medical imaging, including MedImageInsight, an open-source foundation model for general-domain medical imaging that achieved state-of-the-art or human-expert-level performance across a wide range of tasks and modalities.

He is also a co-founder of the International Skin Imaging Collaboration (ISIC) skin cancer challenges, which have drawn more than 114,000 submissions from over 4,000 participants and helped shape how AI systems in dermatology are benchmarked, validated, and integrated into clinical workflows. His work has emphasized not only building high-performing AI systems, but also developing rigorous evaluation methods that reveal hidden failure modes, biases, and deployment risks in real-world settings. His research has also been covered by major media outlets including CNN, CNBC, Forbes, MedGadget, and The Economist.

24,000+ Google Scholar citations
49 h-index
114,000+ ISIC submissions
Clinical AI evaluation

ISIC skin cancer challenges

Co-founded a benchmark ecosystem that has become foundational to dermatology AI research, human-AI studies, and clinically meaningful evaluation protocols.

Explore ISIC
Vision research

Broad impact across computer vision

Contributions across transformers, few-shot learning, multimodal learning, video understanding, visual question answering, explainability, and fairness in modern AI systems.

See publications
Imaging depth

Cardiac MRI, reconstruction, and physiology

Graduate work spanning self-calibrating parallel reconstruction, free-breathing cardiac cine acquisition, ventricular segmentation, and cardiac electrophysiology modeling.

See CV

Editorial and program leadership

  • Area Chair: ICLR 2018, 2019, 2022 (Highlighted), 2023; ICPR 2022; NeurIPS 2022.
  • Senior Program Committee: AAAI 2022.
  • Associate Editor: IEEE Transactions on Multimedia (2021–Present).
  • Associate Editor: IEEE Journal of Biomedical and Health Informatics, Special Issue on Skin Image Analysis.

Community building

  • Challenge Co-Founder / Co-Organizer: International Skin Imaging Collaboration (ISIC) challenges at ISBI 2016–2017 and MICCAI 2018–2020.
  • Workshop Co-Founder / Co-Organizer: ISIC workshops at MICCAI 2018, CVPR 2019–2021, and ECCV 2022.
  • Benchmark / Workshop Co-Organizer: Learning with Limited Labels and related few-shot learning benchmark efforts at CVPR 2020–2021 and ECCV 2022.
  • Mentorship and service: Microsoft Startups Mentor in Machine Learning; former IBM Research PIC Chair for Computer Vision & Multimedia.
  • IBM Outstanding Technical Achievement Award (2018)Image Analysis for Melanoma Detection
  • IBM Eminence and Excellence Award (2018)Trustworthy AI
  • IBM Outstanding Research Accomplishment Award (2019)
  • IBM Research Image Award (2016)For significant contributions to IBM’s public image through work on skin image analysis.
  • IBM Invention Achievement Awards (2016, 2014, 2013)First, second, and third plateaus, for multiple patent filing activities.
  • IBM Research Division Award (2013)For contributions to visual recognition technologies.
  • IBM Eminence and Excellence Award (2012)For efforts in organizing Greater New York Multimedia and Vision Workshop.
2026 · International Journal of Computer Vision

Generative Enhancement for 3D Medical Images

Generative modeling for synthesizing and enhancing 3D medical images with strong volumetric consistency and controllable generation.

Read paper
2025 · Machine Learning for Health

CancerGUIDE: Cancer Guideline Understanding via Internal Disagreement Estimation

Guideline-grounded cancer treatment reasoning with internal disagreement estimation for more reliable clinical decision support.

Read paper
2025 · npj Digital Medicine

Automated triage of cancer-suspicious skin lesions with 3D total-body photography

Large-scale dermatology AI for triage using 3D total-body photography in clinically realistic screening workflows.

Read article
2025 · Journal of Imaging Informatics in Medicine

From Embeddings to Accuracy: Comparing Foundation Models for Radiographic Classification

Real-world evaluation of radiographic foundation models showing strong downstream performance from MedImageInsight.

Read article
2025 · Medical Imaging with Deep Learning

Beyond the Prompt: Deploying Medical Foundation Models on Diverse Chest X-ray Populations

Deployment-focused analysis of how medical foundation models behave across heterogeneous chest X-ray populations and acquisition settings.

Read paper
2025 · Nature Machine Intelligence

Exploring Scalable Medical Image Encoders Beyond Text Supervision

Scalable representation learning for medical imaging that extends beyond text-only supervision.

Read article
2024 · ECCV

Fully Authentic Visual Question Answering Dataset from Online Communities

A dataset and benchmark built from authentic community question-answering use cases for long-form visual question answering.

Read paper
2024 · arXiv / Microsoft Research

MedImageInsight: An Open-Source Embedding Model for General Domain Medical Imaging

Open-source medical imaging foundation model spanning diverse modalities and downstream tasks.

Read paper
2024 · Journal of Investigative Dermatology

Expert Agreement on the Presence and Spatial Localization of Melanocytic Features in Dermoscopy

Careful characterization of expert agreement for melanocytic dermoscopic features and their spatial localization.

Read article
2023 · Nature Medicine

A reinforcement learning model for AI-based decision support in skin cancer

Reinforcement-learning-based decision support for dermatology AI in clinically grounded settings.

Read article

Instruction and guest lecturing

  • Columbia University: Guest Lecturer in Computer Vision (2018).
  • NYU Tandon School of Engineering: Guest Lecturer in Computer Vision (2016), including medical imaging topics.
  • Stevens Institute of Technology: Adjunct Professor in Artificial Intelligence (2014–2016).

Mentorship

Across research roles in industry, Noel has contributed to mentoring interns, collaborators, and early-career researchers working in computer vision, multimodal learning, and medical AI.

This mentorship has complemented his formal teaching activity by helping develop research direction, experimental rigor, and publication-quality work in both foundational vision research and clinically grounded machine learning.

Selected patents

Featured coverage

  • CNBC [Archived] - Microsoft announces new AI tools to help ease workload for doctors and nurses
  • Forbes [Archived] - Microsoft Announces Numerous New AI Tools Dedicated To Healthcare
  • The Economist - Huge foundation models are turbo-charging AI progress
  • CNN [Archived] - IBM uses a smartphone to help diagnose skin cancer
  • MedGadget [Archived] - Interview on AI for skin cancer diagnosis
  • VentureBeat - Skin cancer meets its worst nightmare: IBM