Deep Learning for Biomedical Imaging
Medical image analysis with deep learning: image processing, neural networks, segmentation, transformers, generative AI, and clinical applications.
- Course
- BMD-ENG-495
- Term
- Winter 2025
- Offering
- Archived course
Overview
This page preserves the Winter 2025 offering and its teaching materials. See the course catalog for current availability.
Medical image analysis helps researchers and clinicians understand, visualize, and quantify images. This course introduces deep learning methods for analyzing X-rays, CT scans, and MRI scans.
Offered by the Department of Biomedical Engineering with the Machine and Hybrid Intelligence Lab of Radiology, the course covers image processing, deep learning fundamentals, and applications in medical image analysis.
Learning objectives
- Understand the basic concepts of medical image analysis and deep learning.
- Develop practical skills for using deep learning techniques for medical image analysis.
- Understand how deep learning algorithms can be applied to real-world medical image analysis problems.
- Be able to implement deep learning algorithms using open-source libraries such as PyTorch.
- Learn how to evaluate and validate the performance of deep learning algorithms in medical image analysis.
Syllabus & materials
Lecture notes and presentation slides are the primary course materials.
Lectures 1 & 2
Introduction to Medical Imaging & Analysis
- Overview of medical image analysis and its importance in healthcare
- Types of medical images (X-rays, CT scans, MRI scans, etc.)
- Challenges in medical image analysis
- Overview of deep learning and its applications in medical image analysis
Lectures 3 & 4
Image Processing Techniques
- Image preprocessing (noise reduction, normalization, etc.) (Lecture 3)
- Image segmentation (thresholding, clustering, region growing, etc.) (Lecture 4)
- Feature extraction (texture analysis, radiomics, etc.) (Lecture 4)
Lecture 5
Lecture 6
Modern CNNs and Recurrent Neural Networks (RNNs)
Lecture 7
Object Detection and Image Segmentation with DL
- RCNN, Fast-RCNN, Yolo, SegNet, FCN (Tiramisu), U-Net, etc.
Lecture 8
Transformers / ViT
- Self-attention mechanism
- Multi-head attention
- Other Transformer architectures (e.g., BERT, GPT)
- ViT
Bin Wang · Guest lecturer
Lecture 9
Evaluating DL Algorithms and Data Augmentation
Lecture 10
Lecture 11
Radiology Imaging and Analysis from a Radiologist’s Perspective
Dr. Gorkem Durak · Invited speaker
No material was linked in the archived syllabus.
Lecture 12
Diffusion Generative AI for Colonoscopy/Endoscopy Applications
Dr. Vanshali Sharma · Invited speaker
Lecture 13
Lecture 14
Self-Supervised Learning for Medical Imaging
Reading list
Image Processing, Analysis, and Machine Vision
Image Processing, Analysis, and Machine Vision. M. Sonka, V., Hlavac, R. Boyle. Nelson Engineering, 2014.
Optional readingMedical Imaging Signals and Systems
Medical Imaging Signals and Systems, Jerry Prince & Jonathan Links, Publisher: Prentice Hall.
Optional readingDeep Learning
Deep Learning, Goodfellow Ian, Bengio Yoshua, and Courville Aaron, 2016, Freely available, MIT Press.
Optional readingNotes on the archived syllabus
- The source heading identifies Winter 2025; the source illustration is labeled Winter 2024.
- The source section labeled Lectures 5-to-9 also contains lectures 10–14. All fourteen lecture numbers are preserved.
- The source does not identify a lead instructor, meeting dates, location, grading policy or prerequisites.
- Lecture 11 has an invited speaker but no linked material in the source.
