Summary of Addressing Small and Imbalanced Medical Image Datasets Using Generative Models: a Comparative Study Of Ddpm and Pggans with Random and Greedy K Sampling, by Iman Khazrak et al.
Addressing Small and Imbalanced Medical Image Datasets Using Generative Models: A Comparative Study of DDPM and PGGANs with Random and Greedy K Sampling
by Iman Khazrak, Shakhnoza Takhirova, Mostafa M. Rezaee, Mehrdad Yadollahi, Robert C. Green II, Shuteng Niu
First submitted to arxiv on: 17 Dec 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper investigates the application of generative models to address data scarcity and imbalance issues in medical image classification. The authors explore two models, Denoising Diffusion Probabilistic Models (DDPM) and Progressive Growing Generative Adversarial Networks (PGGANs), for generating synthetic images that augment small, imbalanced datasets. Four different models are tested using these synthetic images: a custom CNN, Untrained VGG16, Pretrained VGG16, and Pretrained ResNet50. The authors evaluate the performance of these models using classification metrics and Frechet Inception Distance (FID) to assess the quality of the generated images. The results show that DDPM consistently generates more realistic images with lower FID scores and improves classification metrics across all models and datasets. This study demonstrates the effectiveness of DDPM in enhancing model robustness, stability, and accuracy. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper helps us make better medical image classification models by creating more training data. Medical imaging is a very important field where doctors use pictures to diagnose diseases. But sometimes we don’t have enough good-quality images to train our models. This problem makes it hard for our models to be accurate and reliable. The authors of this paper think that by generating new, realistic images using special computer algorithms (called generative models), we can help solve this problem. They test these algorithms on different kinds of medical image classification models and show that one algorithm in particular, called Denoising Diffusion Probabilistic Models (DDPM), works really well. This means that DDPM can be used to make our medical image classification models more accurate and reliable. |
Keywords
» Artificial intelligence » Classification » Cnn » Diffusion » Image classification
