Please use this identifier to cite or link to this item:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/74436
Title: | 以深度學習分類皮膚癌光學同調斷層掃描影像 Deep Learning Based Classification of Skin Cancer Optical Coherence Tomography Images |
Authors: | Ting-An Kuo 郭庭安 |
Advisor: | 曾雪峰(Snow H. Tseng) |
Keyword: | 深度學習,卷積神經網路,光學同調斷層掃描,皮膚癌,電腦輔助診斷, Deep learning,CNN,OCT,CAD,Skin cancer,SCC, |
Publication Year : | 2019 |
Degree: | 碩士 |
Abstract: | 在本研究中,我們建立了一個電腦輔助診斷(CAD)工具,提出一種結合深度學習與全域式光學同調斷層掃描(Full-Field Optical Coherence Tomography, FF-OCT)數據中自動診斷皮膚癌之有效方法。本論文採用卷積神經網路(Convolutional Neural Network, CNN) 架構訓練可提取皮膚特徵之分類器以加速檢測鱗狀細胞癌(Squamous Cell Carcinoma, SCC),並利用集成學習提升診斷能力,降低誤判程度。此外,我們已經定量地顯示,本論文所提出之方法經過10次交叉驗證於測試集上獲得78.89%之準確率。結果表明,本研究方法於OCT皮膚癌診斷之可行性。 Here we report an effective approach to automatically diagnose skin cancer in Full-Field Optical Coherence Tomography (FF-OCT). We employ a Convolutional Neural Network (CNN) framework to speed up the detection of Squamous Cell Carcinoma (SCC). In addition, our trained neural network yielded performance of 78.89% accuracy by 10-fold cross-validation. Experimental results demonstrate the effectiveness of the proposed method for the automated diagnosis of skin cancer. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/74436 |
DOI: | 10.6342/NTU201902797 |
Fulltext Rights: | 有償授權 |
Appears in Collections: | 光電工程學研究所 |
Files in This Item:
File | Size | Format | |
---|---|---|---|
ntu-108-1.pdf Restricted Access | 4.49 MB | Adobe PDF |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.