Completed
Demonstrate feasibility of applying the propose to other imaging modalities and patient cohort
Class Central Classrooms beta
YouTube videos curated by Class Central.
Classroom Contents
Towards Generalist Imaging Using Multimodal Self-Supervised Learning - Mars Huang
Automatically move to the next video in the Classroom when playback concludes
- 1 Intro
- 2 Increase of Medical Imaging Utilization Can Hurt Patient
- 3 Limitation 1: Supervised learning requires large sc labeled datasets
- 4 Limitation 2: Few Medical Imaging Models Consider Clinical Context
- 5 Prototyping Methods Using Cohort of Pulmonary Embolism Patients
- 6 Specific Aims
- 7 Challenges For Pulmonary Embolism Detection
- 8 PENet
- 9 Fusion Types
- 10 Major types of self-supervised method for images
- 11 Learning global representations can be limiting
- 12 Global & Local Representations for Images using Attention G
- 13 Representation Learning Objective
- 14 Retrieval Results
- 15 Fine-tune Classification
- 16 Strategies for Generating Class Prompts
- 17 Zero-shot Classification Results
- 18 Next Steps
- 19 Generalizability to Other Downstream Tasks
- 20 Demonstrate feasibility of applying the propose to other imaging modalities and patient cohort
