ChestX-ray

From DIR
Jump to: navigation, search
Type declaration dctypes:Dataset
Title ChestX-ray
Description The chest X-ray is one of the most commonly accessible radiological examinations for screening and diagnosisof many lung diseases. A tremendous number of X-rayimaging studies accompanied by radiological reports areaccumulated and stored in many modern hospitals’ Pic-ture Archiving and Communication Systems (PACS). Onthe other side, it is still an open question how this typeof hospital-size knowledge database containing invaluableimaging informatics (i.e., loosely labeled) can be used to fa-cilitate the data-hungry deep learning paradigms in build-ing truly large-scale high precision computer-aided diagno-sis (CAD) systems. chest X-ray database,namely “ChestX-ray8”, which comprises 108,948 frontal-view X-ray images of 32,717 unique patients with the text-mined eight disease image labels (where each image canhave multi-labels), from the associated radiological reportsusing natural language processing. Importantly, we demon-strate that these commonly occurring thoracic diseases canbe detected and even spatially-located via a unified weakly-supervised multi-label image classification and disease lo-calization framework, which is validated using our proposeddataset. Although the initial quantitative results are promis-ing as reported, deep convolutional neural network based“reading chest X-rays” (i.e., recognizing and locating thecommon disease patterns trained with only image-level la-bels) remains a strenuous task for fully-automated high pre-cision CAD systems.

ChestX-ray14: Hospital-scale Chest X-ray Database and Benchmarks on Weakly-Supervised Classification and Localization of Common Thorax Diseases (PDF Download Available). Available from: https://www.researchgate.net/publication/320068322_ChestX-ray14_Hospital-scale_Chest_X-ray_Database_and_Benchmarks_on_Weakly-Supervised_Classification_and_Localization_of_Common_Thorax_Diseases [accessed Mar 25 2018].
Publisher https://www.nih.gov/
HTML page https://www.nih.gov/news-events/news-releases/nih-clinical-center-provides-one-largest-publicly-available-chest-x-ray-datasets-scientific-community
Update frequency Quaterly
Alternative titles NIH Chest X-ray Dataset, ChestX-ray8, ChestX-ray14, CXR8
References https://www.nih.gov/news-events/news-releases/nih-clinical-center-provides-one-largest-publicly-available-chest-x-ray-datasets-scientific-community
Citations https://www.researchgate.net/publication/320068322_ChestX-ray14_Hospital-scale_Chest_X-ray_Database_and_Benchmarks_on_Weakly-Supervised_Classification_and_Localization_of_Common_Thorax_Diseases
Related material https://www.kaggle.com/nih-chest-xrays/sample, http://academictorrents.com/details/557481faacd824c83fbf57dcf7b6da9383b3235a/tech
Preferred prefix ChestX-ray
Version linking ChestX-ray14
Publications using this dataset http://lanl.arxiv.org/pdf/1711.05225, http://openaccess.thecvf.com/content_cvpr_2017/papers/Wang_ChestX-ray8_Hospital-Scale_Chest_CVPR_2017_paper.pdf
Publication number 2
Geographic area All over the world
Method entities in publications Deep Learning, Bootstrap, Classification Model
Top methods in publications Deep Learning (100.00%), Classification Model (50.00%), Bootstrap(50.00%)
Availability Publicly available
Facts about "ChestX-ray"
AvailabilityPublicly available +
Cito:citesAsAuthorityhttps://www.researchgate.net/publication/320068322_ChestX-ray14_Hospital-scale_Chest_X-ray_Database_and_Benchmarks_on_Weakly-Supervised_Classification_and_Localization_of_Common_Thorax_Diseases +
Cito:citesAsDataSourcehttp://lanl.arxiv.org/pdf/1711.05225 + and http://openaccess.thecvf.com/content_cvpr_2017/papers/Wang_ChestX-ray8_Hospital-Scale_Chest_CVPR_2017_paper.pdf +
Dct:accrualPeriodicityQuaterly +
Dct:alternativeNIH Chest X-ray Dataset +, ChestX-ray8 +, ChestX-ray14 + and CXR8 +
Dct:descriptionThe chest X-ray is one of the most commonl
The chest X-ray is one of the most commonly accessible radiological examinations for screening and diagnosisof many lung diseases. A tremendous number of X-rayimaging studies accompanied by radiological reports areaccumulated and stored in many modern hospitals’ Pic-ture Archiving and Communication Systems (PACS). Onthe other side, it is still an open question how this typeof hospital-size knowledge database containing invaluableimaging informatics (i.e., loosely labeled) can be used to fa-cilitate the data-hungry deep learning paradigms in build-ing truly large-scale high precision computer-aided diagno-sis (CAD) systems. chest X-ray database,namely “ChestX-ray8”, which comprises 108,948 frontal-view X-ray images of 32,717 unique patients with the text-mined eight disease image labels (where each image canhave multi-labels), from the associated radiological reportsusing natural language processing. Importantly, we demon-strate that these commonly occurring thoracic diseases canbe detected and even spatially-located via a unified weakly-supervised multi-label image classification and disease lo-calization framework, which is validated using our proposeddataset. Although the initial quantitative results are promis-ing as reported, deep convolutional neural network based“reading chest X-rays” (i.e., recognizing and locating thecommon disease patterns trained with only image-level la-bels) remains a strenuous task for fully-automated high pre-cision CAD systems.

ChestX-ray14: Hospital-scale Chest X-ray Database and Benchmarks on Weakly-Supervised Classification and Localization of Common Thorax Diseases (PDF Download Available). Available from: https://www.researchgate.net/publication/320068322_ChestX-ray14_Hospital-scale_Chest_X-ray_Database_and_Benchmarks_on_Weakly-Supervised_Classification_and_Localization_of_Common_Thorax_Diseases [accessed Mar 25 2018].
on_Thorax_Diseases [accessed Mar 25 2018]. +
Dct:publisherhttps://www.nih.gov/ +
Dct:referenceshttps://www.nih.gov/news-events/news-releases/nih-clinical-center-provides-one-largest-publicly-available-chest-x-ray-datasets-scientific-community +
Dct:titleChestX-ray +
Foaf:pagehttps://www.nih.gov/news-events/news-releases/nih-clinical-center-provides-one-largest-publicly-available-chest-x-ray-datasets-scientific-community +
Geographic areaAll over the world +
Idot:preferredPrefixChestX-ray +
Methods in publicationsDeep Learning +, Bootstrap + and Classification Model +
Pav:hasCurrentVersionChestX-ray14 +
Publication number2 +
Rdf:typedctypes:Dataset +
Rdfs:seeAlsohttps://www.kaggle.com/nih-chest-xrays/sample + and http://academictorrents.com/details/557481faacd824c83fbf57dcf7b6da9383b3235a/tech +
Top methods in publicationsDeep Learning (100.00%), Classification Model (50.00%), Bootstrap(50.00%) +