{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/conditional-infilling-gans-for-data","title":"Conditional Infilling GANs for Data Augmentation in Mammogram Classification","arxiv_id":"1807.08093","date":"2018-07-21","proceeding":null,"authors":["Eric Wu","Kevin Wu","David Cox","William Lotter"],"abstract":"Deep learning approaches to breast cancer detection in mammograms have\nrecently shown promising results. However, such models are constrained by the\nlimited size of publicly available mammography datasets, in large part due to\nprivacy concerns and the high cost of generating expert annotations. Limited\ndataset size is further exacerbated by substantial class imbalance since\n\"normal\" images dramatically outnumber those with findings. Given the rapid\nprogress of generative models in synthesizing realistic images, and the known\neffectiveness of simple data augmentation techniques (e.g. horizontal\nflipping), we ask if it is possible to synthetically augment mammogram datasets\nusing generative adversarial networks (GANs). We train a class-conditional GAN\nto perform contextual in-filling, which we then use to synthesize lesions onto\nhealthy screening mammograms. First, we show that GANs are capable of\ngenerating high-resolution synthetic mammogram patches. Next, we experimentally\nevaluate using the augmented dataset to improve breast cancer classification\nperformance. We observe that a ResNet-50 classifier trained with GAN-augmented\ntraining data produces a higher AUROC compared to the same model trained only\non traditionally augmented data, demonstrating the potential of our approach.","url_abs":"http://arxiv.org/abs/1807.08093v2","url_pdf":"http://arxiv.org/pdf/1807.08093v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"conditional-infilling-gans-for-data","repo_url":"https://github.com/ericwu09/mammo-cigan","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"breast-cancer-detection","task_name":"Breast Cancer Detection"},{"task_slug":"cancer-classification","task_name":"Cancer Classification"},{"task_slug":"classification-1","task_name":"Classification"},{"task_slug":"data-augmentation","task_name":"Data Augmentation"},{"task_slug":"classification","task_name":"General Classification"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1807.08093","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}