Managing Medical Imaging Datasets: From Curation to Evaluation
High-quality data is the backbone of impactful medical AI. This workshop explores the full lifecycle of medical imaging datasets—from smart curation strategies and synthetic data generation to robust evaluation techniques. You will gain exposure to cutting-edge research and hands-on tools that are reshaping how imaging data is managed in medical AI workflows.
In the first half, leading researchers and practitioners will share best practices in dataset quality control, synthetic data generation, and reproducible evaluation. In the second half, participants will take part in a hands-on tutorial using FiftyOne, a powerful open-source toolkit, to explore and curate datasets across modalities such as X-ray, MRI, CT, and ultrasound.
Whether you are a researcher looking to improve dataset integrity, a clinician collaborating on AI development, or an ML engineer scaling healthcare models, this session offers practical insights and tools to level up your work with medical imaging data.
Target Audience:
- Medical imaging researchers
- Clinical collaborators in AI projects
- ML engineers and data scientists in healthcare
- Professionals involved in dataset annotation and evaluation
Tentative Program:
8:30 AM – 8:45 AM | Welcome and Introduction
Overview of workshop goals, structure, and the importance of robust data practices in medical imaging.
8:45 – 10:15 AM | Talk Session: Innovations in Dataset Management
Speaker: TBD
A high-level perspective on data challenges in medical imaging and the growing role of curated and synthetic datasets.
Short talks + Q&A from leading researchers and practitioners.
10:15 AM – 10:30 AM | Coffee Break & Networking
10:30 AM – 12:15 PM | Hands-On Tutorial: Managing Medical Imaging Datasets with FiftyOne
Facilitators: TBD
12:15 PM – 12:30 PM | Wrap-Up and Discussion
Recap, Q&A, and sharing of additional resources for continued learning.