Summary of Longitudinal Mammogram Exam-based Breast Cancer Diagnosis Models: Vulnerability to Adversarial Attacks, by Zhengbo Zhou et al.
Longitudinal Mammogram Exam-based Breast Cancer Diagnosis Models: Vulnerability to Adversarial Attacks
by Zhengbo Zhou, Degan Hao, Dooman Arefan, Margarita Zuley, Jules Sumkin, Shandong Wu
First submitted to arxiv on: 29 Oct 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 The paper proposes a novel attack method for longitudinal breast cancer detection models that detects temporal imaging feature changes between sequential mammogram exams. The attack exploits the relationship between two sequential images to achieve significant effectiveness, outperforming state-of-the-art attacks in fooling diagnosis models. The proposed method is implemented using black-box attacking and shows resilience even when defending methods like adversarial training are used. This study demonstrates the importance of evaluating the robustness of longitudinal models against adversarial attacks. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Breast cancer detection uses mammogram images to diagnose breast cancer. A new attack can trick these models into giving wrong answers by looking at how two pictures taken together look different. This attack is really good and works even when doctors try to stop it. The study tested the attack on 590 women with breast cancer and found that it worked better than other attacks. |
