coursera-ai4med

This is main repository to group the my course work as part of Deeplearning.ai - AI for Medicine Specialization by deeplearning.ai.

Instructors

• Pranav Rajpurkar, Instructor, PhD Candidate Stanford University
• Bora Uyumazturk
• Amirhossein Kiani
• Eddy Shyu

Specialization

• Course 1: AI for Medical Diagnosis
• Course 2: AI for Medical Prognosis
• Course 3: AI for Medical Treatment

Eureka! moment

If you plan to only audit class, one thing to focus on

• Although radiology image classification and segmentation using a CNN in TensorFlow is interesting, survival analysis using different traditional ML methods are a nice comparison and contrast between modern and classical AI, ML and DL approaches.
• GradCAM is a novel approach to overlay heatmaps on radiology images.

Key Takeaways

• In Healthcare, there are common data challenges of: data access, data size, and data quality.
• Validation need clear methods, evaluation metrics and explanation.
• Despite the hype of AI and deep learning (DL), traditionl ML methods (e.g, decision trees, random forest) still very effective in real world scenarios.

Course work

• coursera-ai4med-course1
• coursera-ai4med-course2
• coursera-ai4med-course3

Related

• Coursera, Deeplearning.ai: Deep Learning Specialization
• Coursera, Machine Learning (Ng, Stanford)