{"id":602,"date":"2016-07-01T12:35:01","date_gmt":"2016-07-01T17:35:01","guid":{"rendered":"https:\/\/my.dev.vanderbilt.edu\/masi\/?p=602"},"modified":"2017-02-08T14:16:31","modified_gmt":"2017-02-08T19:16:31","slug":"deep-learning-for-brain-tumor-classification","status":"publish","type":"post","link":"https:\/\/my.dev.vanderbilt.edu\/masi\/2016\/07\/deep-learning-for-brain-tumor-classification\/","title":{"rendered":"Deep Learning for Brain Tumor Classification"},"content":{"rendered":"<p>Justin S. Paul, Andrew J. Plassard, Bennett A. Landman, Daniel Fabbri. \u201cDeep Learning for Brain Tumor Classification.\u201d In Proceedings of the SPIE Medical Imaging Conference. Orlando, Florida, February 2017. Oral presentation.<\/p>\n<h2>Abstract<\/h2>\n<p>Recent research has shown that deep learning methods have performed well on supervised machine learning,\u00a0image classification tasks. The purpose of this study is to apply deep learning methods to classify brain images\u00a0with different tumor types: meningioma, glioma, and pituitary. A dataset was publicly released containing 3,064\u00a0T1-weighted contrast enhanced MRI (CE-MRI) brain images from 233 patients with either meningioma, glioma,\u00a0or pituitary tumors split across axial, coronal, or sagittal planes. This research focuses on the 989 axial images\u00a0from 191 patients in order to avoid confusing the neural networks with three dierent planes containing the same\u00a0diagnosis. Two types of neural networks were used in classication: fully connected and convolutional neural\u00a0networks. Within these two categories, further tests were computed via the augmentation of the original 512&#215;512\u00a0axial images. Through rotating, shifting, and mirroring, data sizes could be increased with the sacrice of image\u00a0clarity. Training neural networks over the axial data has proven to be accurate in its classications with an\u00a0average ve-fold cross validation of 91.43% on the best trained neural network. This result demonstrates that a\u00a0more general method (i.e. deep learning) can outperform specialized methods that require image dilation and\u00a0ring-forming subregions on tumors.<\/p>\n<figure id=\"attachment_1099\" aria-describedby=\"caption-attachment-1099\" style=\"width: 640px\" class=\"wp-caption alignnone\"><img loading=\"lazy\" decoding=\"async\" class=\"size-full wp-image-1099\" src=\"https:\/\/my.dev.vanderbilt.edu\/masi\/wp-content\/uploads\/sites\/2304\n2661\/2016\/07\/Screen-Shot-2017-02-08-at-1.15.49-PM.png\" alt=\"(A) A standard fully connected neural network where each layer's node is connected to each node from the previous layer.7 (B) A convolutional neural network connecting a covolutional layer to a pooling layer.\" width=\"640\" height=\"534\" \/><figcaption id=\"caption-attachment-1099\" class=\"wp-caption-text\">(A) A standard fully connected neural network where each layer&#8217;s node is connected to each node from the<br \/>previous layer.7 (B) A convolutional neural network connecting a covolutional layer to a pooling layer.<\/figcaption><\/figure>\n","protected":false},"excerpt":{"rendered":"<p>Justin S. Paul, Andrew J. Plassard, Bennett A. Landman, Daniel Fabbri. \u201cDeep Learning for Brain Tumor Classification.\u201d In Proceedings of the SPIE Medical Imaging Conference. Orlando, Florida, February 2017. Oral presentation. Abstract Recent research has shown that deep learning methods have performed well on supervised machine learning,\u00a0image classification tasks. The purpose of this study is&#8230;<\/p>\n","protected":false},"author":6322,"featured_media":1099,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[27,1],"tags":[],"class_list":["post-602","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-big-data","category-news"],"_links":{"self":[{"href":"https:\/\/my.dev.vanderbilt.edu\/masi\/wp-json\/wp\/v2\/posts\/602","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/my.dev.vanderbilt.edu\/masi\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/my.dev.vanderbilt.edu\/masi\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/my.dev.vanderbilt.edu\/masi\/wp-json\/wp\/v2\/users\/6322"}],"replies":[{"embeddable":true,"href":"https:\/\/my.dev.vanderbilt.edu\/masi\/wp-json\/wp\/v2\/comments?post=602"}],"version-history":[{"count":4,"href":"https:\/\/my.dev.vanderbilt.edu\/masi\/wp-json\/wp\/v2\/posts\/602\/revisions"}],"predecessor-version":[{"id":1100,"href":"https:\/\/my.dev.vanderbilt.edu\/masi\/wp-json\/wp\/v2\/posts\/602\/revisions\/1100"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/my.dev.vanderbilt.edu\/masi\/wp-json\/wp\/v2\/media\/1099"}],"wp:attachment":[{"href":"https:\/\/my.dev.vanderbilt.edu\/masi\/wp-json\/wp\/v2\/media?parent=602"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/my.dev.vanderbilt.edu\/masi\/wp-json\/wp\/v2\/categories?post=602"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/my.dev.vanderbilt.edu\/masi\/wp-json\/wp\/v2\/tags?post=602"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}