Youven Zeghlache
RISE-MICCAI 2025
Past Event
Medical Imaging Deep Learning Tutorial Machine Learning Classification MICCAI Beginner-Friendly

Overview

Join us for a comprehensive hands-on tutorial at RISE-MICCAI 2025, designed specifically for beginners in medical image classification! As part of the inaugural edition of RISE-MICCAI, and in my role as tutorial coordinator of RISE, I’m launching this new educational initiative to bridge the knowledge gap for newcomers to the field.

Event Details

  • Date: Saturday, April 19th, 2025
  • Time: 2:00 PM CET
  • Event: RISE-MICCAI 2025 Tutorial Session
  • Format: Hands-on Tutorial
  • Target Audience: Students, healthcare innovators, curious developers, and anyone new to medical imaging

What You’ll Learn

This comprehensive tutorial will equip you with essential tools, insights, and practical tips to confidently approach medical imaging projects:

🔧 Data Handling Fundamentals

  • Medical Data Preprocessing: Proper techniques for cleaning and preparing medical images
  • Labeling Best Practices: How to handle annotations and ground truth data effectively
  • Dataset Splitting: Right and wrong ways to divide your data for training, validation, and testing

⚠️ Common Pitfalls to Avoid

  • Overfitting Prevention: Recognizing and mitigating overfitting in medical imaging contexts
  • Data Leakage: Understanding and preventing subtle forms of data contamination
  • Validation Mistakes: Ensuring your model evaluation is truly representative

📊 Evaluation Metrics Mastery

  • AUROC (Area Under ROC Curve): When and how to use it properly
  • Sensitivity and Specificity: Critical metrics for medical applications
  • Precision and Recall: Understanding their importance in healthcare contexts
  • Choosing the Right Metrics: Matching evaluation criteria to your specific medical task

⚖️ Handling Challenging Scenarios

  • Class Imbalance: Effective strategies for dealing with unequal class distributions
  • Rare Disease Cases: Special considerations for uncommon conditions
  • Small Dataset Challenges: Making the most of limited medical data

🔍 Model Transparency and Trust

  • Explainability Techniques: Tools and methods to understand model decisions
  • Clinical Interpretability: Making AI decisions understandable to healthcare professionals
  • Building Trust: Ensuring your models are reliable and trustworthy in medical settings

Why This Tutorial Matters

Medical image classification is a rapidly growing field with immense potential to improve patient care. However, the unique challenges of medical data require specialized knowledge and careful attention to detail. This tutorial addresses the critical gap between general machine learning education and the specific requirements of medical imaging applications.

Target Audience

This tutorial is perfect for:

  • Students entering the medical AI field
  • Healthcare Innovators looking to understand AI capabilities
  • Software Developers transitioning to medical applications
  • Researchers new to medical imaging
  • Anyone curious about the intersection of AI and healthcare

Tutorial Format

The session will be highly interactive and practical, featuring:

  • Live demonstrations of common mistakes and their solutions
  • Hands-on exercises with real medical imaging scenarios
  • Q&A sessions to address specific challenges
  • Practical tips that can be immediately applied to your projects

About RISE-MICCAI 2025

This tutorial is part of the inaugural RISE-MICCAI conference, a new initiative focused on making medical imaging AI more accessible and inclusive. As the tutorial coordinator for RISE, I’m committed to providing high-quality educational content that empowers newcomers to succeed in this exciting field.

Key Takeaways

By the end of this tutorial, you’ll have:

  • A solid foundation in medical image classification best practices
  • Practical knowledge of common pitfalls and how to avoid them
  • Confidence to start your own medical imaging projects
  • Access to tools and resources for continued learning
  • A network of peers and experts in the field

Resources and Follow-up

Participants will receive:

  • Comprehensive tutorial materials and code examples
  • Recommended reading list for further learning
  • Access to practice datasets
  • Contact information for ongoing support and questions

This tutorial represents my commitment to education and knowledge sharing in the medical AI community. Whether you’re taking your first steps into medical imaging or looking to solidify your foundational knowledge, this session will provide you with the essential tools and confidence to succeed.