CSYE 7374: Applied Deep Learning and Generative Models in Healthcare

Northeastern University · Fall 2025 · Instructor: Matt Ebrahim (Mahmoud Ebrahimkhani)

Course Information

Instructor

Matt Ebrahim (Mahmoud Ebrahimkhani)

Email: m.ebrahimkhani@northeastern.edu

Office Hours: Fridays 11:00 – 12:00 PM ET (Zoom)

Course Details

Meeting Times: Saturdays 12:30 – 3:30 PM ET

Format: Virtual (Zoom)

Course Description

This course provides a comprehensive introduction to generative AI techniques and their applications in healthcare. Students will explore cutting-edge methods spanning from foundational deep learning architectures to state-of-the-art generative models, with hands-on implementation experience throughout.

The course progresses through three main phases:

Foundations & Language Models (Weeks 1-5): Beginning with large language models specialized for biomedical applications (BioBERT, Med-PaLM), students will learn fine-tuning strategies, retrieval-augmented generation (RAG) for clinical knowledge retrieval, and multimodal learning approaches including CLIP and MedCLIP. Self-supervised and contrastive learning methods will provide essential background for understanding modern medical AI systems.

Computer Vision & Generative Models (Weeks 6-10): The course then covers medical image analysis fundamentals including segmentation and classification using CNNs and Vision Transformers, before diving deep into generative approaches. Students will implement GANs for medical image synthesis and master diffusion models—from DDPM theory to practical latent diffusion applications. Special emphasis on synthetic data generation for rare diseases and data augmentation strategies.

Molecular AI & Drug Discovery (Weeks 11-12): Advanced topics include protein language models (AlphaFold, ESM-2) for structure prediction and generative AI methods for drug discovery, including GFlowNets and molecular diffusion models.

Throughout the course, students will complete four hands-on assignments and a final project, gaining practical experience in designing, training, and evaluating generative AI systems for real-world healthcare challenges.

See Complete Syllabus

Course Schedule & Materials

Week Date In‑Class Topic Materials
1 09 / 06 / 25 Introduction & Diffusion‑Model Foundations Slides Notebook
2 09 / 13 / 25 Large Language Models (BioBERT, Med-PaLM, fine-tuning) Slides Notebook
3 09 / 20 / 25 Multimodal Learning (CLIP, MedCLIP, MedGemma) Slides Notebook Homework-01
4 09 / 27 / 25 Variational Autoencoders & Generative Models Slides Notebook
5 10 / 04 / 25 Denoising Diffusion Probabilistic Models Slides Notebook-VAE Notebook-GAN
6 10 / 11 / 25 Intro to Graph Neural Networks Slides Notebook Homework-02
7 10 / 18 / 25 Medical Image Segmentation Slides Notebook
8 10 / 25 / 25 Medical Image Classification Slides Notebook
9 11 / 01 / 25 De Novo Drug Discovery with Transformer Models for SMILES Generation Slides Notebook
10 11 / 08 / 25 Molecular AI Slides
11 11 / 15 / 25 Time Series in Healthcare Slides