本文選題：鉬靶乳腺X線影像　切入點(diǎn)：計(jì)算機(jī)輔助診斷系統(tǒng)　出處：《浙江大學(xué)》2013年博士論文

【摘要】：乳腺癌是全球女性中最為頻發(fā)的惡性腫瘤疾病和癌癥死亡的首要原因。大量研究表明,早期診斷是擴(kuò)大乳腺癌治療方案的選擇空間、降低乳腺癌患者死亡率的關(guān)鍵。鉬靶乳腺X線攝影能顯示觸診無法感知的早期病變,是乳腺癌早期診斷的首選方法。而基于乳腺X線影像的計(jì)算機(jī)輔助診斷(computer-aided diagnosis, CAD)技術(shù)能夠協(xié)助放射科醫(yī)生檢測和分析可疑病灶,提高乳腺癌早期診斷的準(zhǔn)確率。 腫塊是乳腺X線影像中的重要病理征象,其影像表現(xiàn)復(fù)雜多變,容易受到周圍組織的干擾,是乳腺CAD系統(tǒng)的研究重點(diǎn)和難點(diǎn)之一。傳統(tǒng)的乳腺CAD系統(tǒng)以病灶檢測和病理分類為主要功能,其只提示可疑病變而不提供診斷依據(jù)的“黑箱”過程降低了醫(yī)生對(duì)CAD系統(tǒng)的認(rèn)可和信賴。近年來,一種基于圖像內(nèi)容檢索(content-based image retrieval, CBIR)的CAD系統(tǒng)能夠很好的克服傳統(tǒng)CAD系統(tǒng)在輔助診斷方面的不足。 本論文重點(diǎn)探討了基于乳腺X線腫塊影像的CBIR-CAD系統(tǒng)的關(guān)鍵技術(shù),旨在提高腫塊檢測和輔助診斷的性能,為放射科醫(yī)生提供有價(jià)值的“第二參考意見”。本論文的內(nèi)容包括腫塊分割、特征提取和優(yōu)化、腫塊亞型分類以及基于內(nèi)容的相似腫塊檢索。 (1)腫塊分割：提出了兩種基于隨機(jī)游走的自動(dòng)腫塊分割算法。第一種方法利用腫塊等高線圖自動(dòng)標(biāo)記隨機(jī)游走算法所需的種子點(diǎn),將背景種子點(diǎn)設(shè)置為包圍腫塊的閉合輪廓,抑制周圍組織的影響。同時(shí),利用等高線圖的嵌套特性,不斷擴(kuò)張腫塊種子點(diǎn)的范圍,并利用隨機(jī)游走算法獲得一系列分割結(jié)果。最后,定義結(jié)合腫塊尺寸、邊界梯度和灰度信息的評(píng)價(jià)函數(shù)并選擇最終的分割結(jié)果。該方法克服了半自動(dòng)隨機(jī)游走算法的應(yīng)用局限性。第二種算法結(jié)合了隨機(jī)游走和Chan-Vese (CV)活動(dòng)輪廓模型的互補(bǔ)優(yōu)勢,首先利用等高線圖自動(dòng)標(biāo)記種子點(diǎn)并進(jìn)行初始隨機(jī)游走分割。然后,利用隨機(jī)游走獲得的概率矩陣調(diào)制CV模型的能量函數(shù),抑制輪廓泄露現(xiàn)象。同時(shí),在輪廓演化過程中不斷更新概率矩陣,并獲得最終的分割結(jié)果。實(shí)驗(yàn)結(jié)果表明本論文提出的兩種分割方法與幾種現(xiàn)有的分割算法相比,具有更好的分割精度、適應(yīng)性和魯棒性。 (2)特征提取和優(yōu)化：以乳腺影像報(bào)告和數(shù)據(jù)系統(tǒng)(Breast Imaging Reporting and Data System, BI-RADS)為依據(jù),分別提取了腫塊灰度、形狀和邊緣特征,并提出了6維基于腫塊等高線拓?fù)渥兓男绿卣�。然�?綜合五種特征性能評(píng)價(jià)指標(biāo),針對(duì)不同的分類目標(biāo),采用濾波式方法對(duì)上述多維特征進(jìn)行優(yōu)化。實(shí)驗(yàn)對(duì)比了優(yōu)化前后的特征向量和分類性能,結(jié)果顯示本論文提出的6維特征在區(qū)別腫塊的5種邊緣亞型方面表現(xiàn)出顯著的優(yōu)勢。特征優(yōu)化過程在降低特征維度的同時(shí),提高了分類性能和效率。 (3)腫塊亞型分類：提出了一種結(jié)合支持向量機(jī)(support vector machine, SVM)和二叉決策樹(binary decision tree, BDT)的動(dòng)態(tài)多分類方法,對(duì)腫塊的4種形狀亞型和5種邊緣亞型進(jìn)行分類。該方法結(jié)合了BDT較高的運(yùn)算效率和SVM良好的分類性能,同時(shí)能夠根據(jù)查詢腫塊的實(shí)際特征,動(dòng)態(tài)調(diào)整節(jié)點(diǎn)分類器和分類步驟,從而有效抑制傳統(tǒng)SVM-BDT方法的錯(cuò)誤累積問題,在提高分類精度的同時(shí),具有較高的訓(xùn)練和測試效率。 (4)基于內(nèi)容的相似腫塊檢索：首先結(jié)合亞型分類器的概率輸出結(jié)果和特征向量的歐式距離定義相似性測度函數(shù),獲得第一輪按例檢索(query-by-example, QBE)結(jié)果。然后,在相關(guān)反饋(relevant feedback, RFb)環(huán)節(jié),通過交互操作對(duì)第一輪檢索結(jié)果進(jìn)行相關(guān)性標(biāo)記。針對(duì)實(shí)際應(yīng)用中用戶標(biāo)定樣本數(shù)量過少的問題,利用未標(biāo)記樣本與標(biāo)記樣本的核空間歐式距離進(jìn)一步擴(kuò)充了反饋樣本的數(shù)量,并訓(xùn)練新的SVM相似性判斷模型。最后,采用留一校驗(yàn)法(leave-one-out cross validation, LOOCV)和準(zhǔn)確率召回率曲線(precision-recall curve, PRC)評(píng)價(jià)了QBE環(huán)節(jié)和RFb環(huán)節(jié)的檢索性能。實(shí)驗(yàn)結(jié)果顯示, QBE環(huán)節(jié)結(jié)合亞型分類器概率輸出和歐式距離的相似性測度函數(shù)比單獨(dú)使用歐式距離具有更好的檢索性能。同時(shí),在用戶標(biāo)定樣本數(shù)和反饋次數(shù)一致的條件下,本論文的RFb環(huán)節(jié)對(duì)系統(tǒng)檢索性能有更大的提升。
[Abstract]:Breast cancer is one of the most frequent malignant tumor diseases and cancer deaths among women in the world . A large number of studies have shown that early diagnosis is the key to expanding breast cancer treatment options and reducing the mortality of breast cancer patients . Mammography of mammography can show the early diagnosis of breast cancer . The computer - aided diagnosis ( CAD ) technique based on breast X - ray image can help the radiologist to detect and analyze suspicious lesions and improve the accuracy of early diagnosis of breast cancer .

In recent years , a CAD system based on content - based image retrieval ( CBIR ) can overcome the shortage of traditional CAD system in auxiliary diagnosis .

This paper focuses on the key techniques of CBIR - CAD system based on breast X - ray mass image , aiming at improving the performance of tumor detection and auxiliary diagnosis , providing valuable " second reference opinion " for radiologists . The contents of this paper include tumor segmentation , feature extraction and optimization , classification of tumor subtypes and retrieval of similar masses based on content .

( 1 ) Tumor segmentation : Two kinds of automatic mass segmentation algorithms based on random walk are presented . The first method uses the contour map to automatically mark the seed point needed by the random walk algorithm , and then sets the background seed point to surround the closed contour of the mass , and then obtains a series of segmentation results .

( 2 ) Feature extraction and optimization : Based on Breast Imaging Reporting and Data System ( BI - RADS ) , the gray scale , shape and edge characteristics of the masses are extracted respectively .

( 3 ) Classification of tumor subtypes : A dynamic multi - classification method combined with support vector machine ( SVM ) and binary decision tree ( BDT ) is proposed to classify the four types of tumor and five types of edge subtypes . The method combines the high operational efficiency of BDT and good classification performance of SVM , and can dynamically adjust the node classifier and classification step according to the actual characteristics of the masses , thereby effectively inhibiting the error accumulation problem of the traditional SVM - BDT method , and has higher training and testing efficiency while improving the classification accuracy .

( 4 ) Based on the similarity measure function of the content - based similarity measure , the results of query - by - example ( QBE ) are firstly combined with the probability output result of the subtype classifier and the Euclidean distance of the feature vector . Then , the retrieval performance of the QBE segment and RFb link is further expanded by interaction operation in the link of relevant feedback ( RFb ) . The results show that the similarity measure function of the probability output and the Euclidean distance of the combined subtype classifier in QBE segment has better retrieval performance than the Euclidean distance alone . At the same time , the RFb link of this paper has a greater improvement in the retrieval performance of the system under the condition of consistent user calibration sample number and feedback frequency .

【學(xué)位授予單位】：浙江大學(xué)
【學(xué)位級(jí)別】：博士
【學(xué)位授予年份】：2013
【分類號(hào)】：R816.4;R737.9

【參考文獻(xiàn)】

相關(guān)期刊論文 前6條

1 沈曄;李敏丹;夏順仁;;基于內(nèi)容的醫(yī)學(xué)圖像檢索技術(shù)[J];計(jì)算機(jī)輔助設(shè)計(jì)與圖形學(xué)學(xué)報(bào);2010年04期

2 郝欣;夏順仁;;一種基于輪廓監(jiān)督的動(dòng)態(tài)規(guī)劃乳腺X線影像腫塊分割算法[J];科技導(dǎo)報(bào);2009年21期

3 孫強(qiáng);;乳腺癌的早期診斷[J];實(shí)用醫(yī)學(xué)雜志;2007年01期

4 顧雅佳;肖勤;;乳腺X線報(bào)告規(guī)范化——BI-RADS介紹[J];中國醫(yī)學(xué)計(jì)算機(jī)成像雜志;2007年05期

5 曹穎;郝欣;朱曉恩;夏順仁;;基于自動(dòng)隨機(jī)游走的乳腺腫塊分割算法[J];浙江大學(xué)學(xué)報(bào)(工學(xué)版);2011年10期

6 郝欣;曹穎;夏順仁;;基于醫(yī)學(xué)圖像內(nèi)容檢索的計(jì)算機(jī)輔助乳腺X線影像診斷技術(shù)[J];中國生物醫(yī)學(xué)工程學(xué)報(bào);2009年06期

相關(guān)博士學(xué)位論文 前1條

1 蘭義華;基于圖像內(nèi)容檢索的乳腺腫塊診斷方法研究[D];華中科技大學(xué);2011年

相關(guān)碩士學(xué)位論文 前1條

1 曹穎;乳腺X線影像腫塊計(jì)算機(jī)輔助診斷算法研究[D];浙江大學(xué);2011年

，

本文編號(hào)：1680526