ChestX-ray8を用いた構造適応型Deep Belief Networkにおける胸部疾患の分類と位置検出の試み

URI http://harp.lib.hiroshima-u.ac.jp/pu-hiroshima/metadata/12596
File
Title
ChestX-ray8を用いた構造適応型Deep Belief Networkにおける胸部疾患の分類と位置検出の試み
Title Alternative
Attempt to Classification and Localization of Thorax Diseases on ChestX-ray8 by Adaptive Structural Learning of Deep Belief Network
Author
氏名 市村 匠
ヨミ イチムラ タクミ
別名 Ichimura Takumi
氏名 鎌田 真
ヨミ カマダ シン
別名 Kamada Shin
Abstract

Abstract—Deep Learning has a hierarchical network architecture to represent the complicated feature of in-put patterns. We have developed the adaptive structure learning method of Deep Belief Network (DBN) that can discover an optimal number of hidden neurons for given input data in a Restricted Boltzmann Machine (RBM) by neuron generation-annihilation algorithm, and hidden layers in DBN. We examined to the learning method to medical open database: CXR8. The CXR8 is one of the most commonly accessible radiological examination for screening and diagnosis of many lung diseases. This paper describes our method accuracy of the classification and localization for the given bounding box(B-Box). The classification ratio for 8 diseases were almost 100% score. A new localization method for DBN is proposed here and the discrete heatmap, the likelihood map of pathologies, was automatically constructed.

Description

開催日:平成30年7月28日
会場:広島工業大学

Journal Title
2018 IEEE SMC Hiroshima Chapter若手研究会講演論文集
Spage
77
Epage
83
Published Date
2018
Publisher
IEEE SMC Hiroshima Chapter
Contributor
市村匠
Language
jpn
NIIType
Conference Paper
Text Version
出版社版
Set
pu-hiroshima