Unsupervised and Semi-Supervised Deep Learning for Medical Imaging
summary
Covers unsupervised and semi-supervised deep learning methods for situations with limited pixel-level labels, focusing on brain tumor MRI segmentation using Stacked Denoising Auto-Encoders (SDAEs) and comparing their performance to fully supervised Convolutional Neural Networks (CNNs). Presents these techniques in a domain-agnostic way so they can be applied beyond medical imaging and highlights recent computer-vision breakthroughs relevant to the approach.
AI-written summary · about
chapters (10)
- 0:00Intro
- 2:04How does cancer accur?
- 4:15Spatting brain cancer
- 8:55Can we use Auto-encoders to extract features?
- 12:03Training Autoencoders on brain MRI
- 15:07Fine-tune the network for classification
- 16:21Segmentation results
- 20:32How good is novelty detection?
- 24:23Fully convolutional ladder networks
- 25:56The future of data-efficient learning
related talks
47mThe whole truth about using deep learning in medical imaging
55mDeep Learning in Medical Image Diagnostics
49mBuilding AI That Learns and Adapts: A Case Study in MRI Diagnostics
25mDeep Learning in Biomedicine: Challenges and Opportunities
43mBuilding AI That Learns and Adapts: A Case Study in MRI Diagnostics
42mBuilding AI That Learns and Adapts: A Case Study in MRI Diagnostics
44mBuilding Adaptive AI for MRI Diagnostics: A Continual Learning Case Study
24mMachine Learning & Medical Imaging: The Future of Early Cancer Detection