{"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/pathological-evidence-exploration-in-deep","title":"Pathological Evidence Exploration in Deep Retinal Image Diagnosis","arxiv_id":"1812.02640","date":"2018-12-06","proceeding":null,"authors":["Yuhao Niu","Lin Gu","Feng Lu","Feifan Lv","Zongji Wang","Imari Sato","Zijian Zhang","Yangyan Xiao","Xunzhang Dai","Tingting Cheng"],"abstract":"Though deep learning has shown successful performance in classifying the\nlabel and severity stage of certain disease, most of them give few evidence on\nhow to make prediction. Here, we propose to exploit the interpretability of\ndeep learning application in medical diagnosis. Inspired by Koch's Postulates,\na well-known strategy in medical research to identify the property of pathogen,\nwe define a pathological descriptor that can be extracted from the activated\nneurons of a diabetic retinopathy detector. To visualize the symptom and\nfeature encoded in this descriptor, we propose a GAN based method to synthesize\npathological retinal image given the descriptor and a binary vessel\nsegmentation. Besides, with this descriptor, we can arbitrarily manipulate the\nposition and quantity of lesions. As verified by a panel of 5 licensed\nophthalmologists, our synthesized images carry the symptoms that are directly\nrelated to diabetic retinopathy diagnosis. The panel survey also shows that our\ngenerated images is both qualitatively and quantitatively superior to existing\nmethods.","url_abs":"http://arxiv.org/abs/1812.02640v1","url_pdf":"http://arxiv.org/pdf/1812.02640v1.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":"pathological-evidence-exploration-in-deep","repo_url":"https://github.com/zzdyyy/Patho-GAN","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":"medical-diagnosis","task_name":"Medical Diagnosis"}],"methods":[{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"interpretability","method_name":"Interpretability"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1812.02640","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}