{"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/chest-x-ray-analysis-of-tuberculosis-by-deep","title":"Chest X-Ray Analysis of Tuberculosis by Deep Learning with Segmentation and Augmentation","arxiv_id":"1803.01199","date":"2018-03-03","proceeding":null,"authors":["Sergii Stirenko","Yuriy Kochura","Oleg Alienin","Oleksandr Rokovyi","Peng Gang","Wei Zeng","Yuri Gordienko"],"abstract":"The results of chest X-ray (CXR) analysis of 2D images to get the\nstatistically reliable predictions (availability of tuberculosis) by\ncomputer-aided diagnosis (CADx) on the basis of deep learning are presented.\nThey demonstrate the efficiency of lung segmentation, lossless and lossy data\naugmentation for CADx of tuberculosis by deep convolutional neural network\n(CNN) applied to the small and not well-balanced dataset even. CNN demonstrates\nability to train (despite overfitting) on the pre-processed dataset obtained\nafter lung segmentation in contrast to the original not-segmented dataset.\nLossless data augmentation of the segmented dataset leads to the lowest\nvalidation loss (without overfitting) and nearly the same accuracy (within the\nlimits of standard deviation) in comparison to the original and other\npre-processed datasets after lossy data augmentation. The additional limited\nlossy data augmentation results in the lower validation loss, but with a\ndecrease of the validation accuracy. In conclusion, besides the more complex\ndeep CNNs and bigger datasets, the better progress of CADx for the small and\nnot well-balanced datasets even could be obtained by better segmentation, data\naugmentation, dataset stratification, and exclusion of non-evident outliers.","url_abs":"http://arxiv.org/abs/1803.01199v1","url_pdf":"http://arxiv.org/pdf/1803.01199v1.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":"chest-x-ray-analysis-of-tuberculosis-by-deep","repo_url":"https://github.com/BrixIA/Brixia-score-COVID-19","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":"data-augmentation","task_name":"Data Augmentation"},{"task_slug":"segmentation","task_name":"Segmentation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1803.01199","atlas_url":"https://app.syntology.ai/?focus=1803.01199","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}