本文關(guān)鍵詞： 深度學(xué)習(xí) 全卷積神經(jīng)網(wǎng)絡(luò) 肝臟分割　出處：《吉林大學(xué)》2017年碩士論文　論文類型：學(xué)位論文

【摘要】：肝臟腫瘤的惡性度高,病情發(fā)展快,死亡率位居世界第二,而我國(guó)是肝癌的高發(fā)地區(qū)之一。隨著醫(yī)學(xué)成像技術(shù)的發(fā)展,計(jì)算機(jī)斷層掃描技術(shù)(CT)在肝臟疾病的相關(guān)診斷中被廣泛應(yīng)用,并且已成為診斷肝臟疾病的首選方法。利用計(jì)算機(jī)圖像處理技術(shù),結(jié)合醫(yī)學(xué)影像診斷技術(shù),對(duì)肝臟疾病進(jìn)行早期診斷、三維建模以及定量分析,能夠使醫(yī)生在術(shù)前掌握充足的數(shù)據(jù),進(jìn)行術(shù)前規(guī)劃,提高手術(shù)的成功率并擬定合理有效的治療方案。從腹部CT影像中準(zhǔn)確可靠地分割肝臟輪廓,是肝臟疾病的早期診斷、肝臟大小及病情的估測(cè)和三維建模的第一步,也是非常關(guān)鍵的一步,其分割結(jié)果對(duì)后續(xù)工作有著直接的影響。在實(shí)際的臨床應(yīng)用中,一般通過(guò)擁有相關(guān)實(shí)踐經(jīng)驗(yàn)和專業(yè)知識(shí)的醫(yī)師從CT影像中手工分割肝臟輪廓,但是,這個(gè)過(guò)程是非常耗費(fèi)時(shí)間和精力的,而且受不同醫(yī)師的主觀因素、經(jīng)驗(yàn)以及知識(shí)的差異的影響,往往會(huì)得到不同的分割結(jié)果。所以,為減輕醫(yī)生的工作負(fù)擔(dān),提高工作效率,也為了獲得更加客觀、準(zhǔn)確的分割結(jié)果,必然要引入計(jì)算機(jī)輔助診斷技術(shù),幫助專業(yè)醫(yī)師分割肝臟CT圖像。傳統(tǒng)的肝臟分割方法,是以圖像處理方法為基礎(chǔ),主要依賴于圖像的一些淺層特征,如:灰度、統(tǒng)計(jì)結(jié)構(gòu)以及紋理等來(lái)分割肝臟輪廓。這種特征可以從圖像中直接獲得,或者通過(guò)人工設(shè)計(jì)的提取算子獲得。這些淺層特征的魯棒性較低,代表性不強(qiáng),也易受到噪聲的干擾。實(shí)踐證明,往往是那些抽象的、深層的特征更具代表性。深度學(xué)習(xí)技術(shù)能夠從大量數(shù)據(jù)中挖掘數(shù)據(jù)深層的抽象的特征,將其應(yīng)用到肝臟分割任務(wù)中能夠提高分割的精度和魯棒性。本文提出了一種基于全卷積神經(jīng)網(wǎng)絡(luò)的CT影像肝臟分割的方法。我們以經(jīng)典的卷積神經(jīng)網(wǎng)絡(luò)Alex Net為基本框架,對(duì)其網(wǎng)絡(luò)布局做了一定的改動(dòng),使其成為全卷積網(wǎng)絡(luò)結(jié)構(gòu)。實(shí)驗(yàn)中使用大量的已標(biāo)注好的腹部CT影像對(duì)網(wǎng)絡(luò)進(jìn)行訓(xùn)練,借由構(gòu)造好的代價(jià)函數(shù)更新網(wǎng)絡(luò)參數(shù)。為解決使用傳統(tǒng)激活函數(shù)造成的梯度消失等問(wèn)題,采用Re LU函數(shù)作為激活函數(shù),同時(shí)為緩解過(guò)擬合問(wèn)題,在網(wǎng)絡(luò)中使用Dropout技術(shù),增加網(wǎng)絡(luò)的泛化能力。由于深度網(wǎng)絡(luò)高層的輸出缺乏足夠的細(xì)節(jié)信息,導(dǎo)致獲得的分割結(jié)果比較粗糙。為解決這個(gè)問(wèn)題,本文提出了一種融合低層和高層特征的網(wǎng)絡(luò)結(jié)構(gòu),通過(guò)將低層的局部細(xì)節(jié)信息與高層的抽象的語(yǔ)義信息融合,進(jìn)而獲得更加精確地分割結(jié)果。實(shí)驗(yàn)結(jié)果表明,本文提出的算法具有較好的魯棒性和精度,同時(shí)與基于Patch的方法相比,本文方法具有更高的效率。
[Abstract]:Liver tumors have a high degree of malignancy, a rapid development of the disease, and a mortality rate of the second highest in the world, and China is one of the regions with a high incidence of liver cancer. With the development of medical imaging technology, Computed tomography (CT) has been widely used in the diagnosis of liver diseases, and has become the first choice in the diagnosis of liver diseases. Early diagnosis, 3D modeling and quantitative analysis of liver diseases enable doctors to master sufficient data and plan before operation. The accurate and reliable segmentation of liver contour from abdominal CT images is the first step in the early diagnosis of liver disease, the estimation of liver size and disease condition, and the three-dimensional modeling. It is also a very critical step. The segmentation results have a direct impact on the follow-up work. In practical clinical applications, the liver contour is usually segmented manually from CT images by doctors with relevant practical experience and professional knowledge, but, This process is very time-consuming and energy-intensive, and often results in different segmentation results due to different doctors' subjective factors, experience and knowledge. In order to obtain more objective and accurate segmentation results, it is necessary to introduce computer-aided diagnosis technology to help doctors to segment liver CT images. Traditional liver segmentation methods are based on image processing methods. It mainly depends on some shallow features of the image, such as grayscale, statistical structure and texture, to segment the liver contour. This feature can be obtained directly from the image. Or it can be obtained by artificial extraction operators. These shallow features are less robust, less representative, and easily disturbed by noise. Practice has proved that these shallow features are often abstract. The deep features are more representative. The deep learning technology can mine the abstract features of the deep layer of the data from a large amount of data. Applying it to liver segmentation can improve the accuracy and robustness of liver segmentation. In this paper, a method of liver segmentation in CT image based on full convolution neural network is proposed. We take the classical convolutional neural network Alex Net as the basic framework. The network layout is modified to make it a full convolution network structure. In the experiment, a large number of tagged abdominal CT images are used to train the network. In order to solve the problem of gradient disappearance caused by the traditional activation function, re LU function is used as the activation function. In order to alleviate the problem of over-fitting, Dropout technology is used in the network. In order to solve this problem, this paper proposes a network structure that combines the features of the lower layer and the higher level, because of the lack of sufficient detail information in the output of the high level network, which results in rough segmentation results. By merging the local details of the lower level with the high-level abstract semantic information, the segmentation results are obtained more accurately. The experimental results show that the proposed algorithm is robust and accurate. At the same time, compared with the method based on Patch, this method has higher efficiency.
【學(xué)位授予單位】：吉林大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：R735.7;TP391.41

【參考文獻(xiàn)】

相關(guān)期刊論文 前3條

1 羅蔓;黃靖;楊豐;;基于多模態(tài)3D-CNNs特征提取的MRI腦腫瘤分割方法[J];科學(xué)技術(shù)與工程;2014年31期

2 劉鑫;陳永健;萬(wàn)洪林;孫娜娜;;基于兩階段區(qū)域生長(zhǎng)的肝內(nèi)血管分割算法[J];計(jì)算機(jī)工程與應(yīng)用;2015年12期

3 廖苗;趙于前;曾業(yè)戰(zhàn);黃忠朝;鄒北驥;;基于圖割和邊緣行進(jìn)的肝臟CT序列圖像分割[J];電子與信息學(xué)報(bào);2016年06期

相關(guān)博士學(xué)位論文 前1條

1 劉宇;基于水平集方法和模糊模型的醫(yī)學(xué)圖像分割算法研究[D];吉林大學(xué);2016年

，

本文編號(hào)：1540222