一种用于前列腺区域分割的改进水平集算法
收稿日期: 2021-01-26
网络出版日期: 2021-08-26
基金资助
国家自然科学基金资助项目(81572530)
An Improved Level Set Algorithm for Prostate Region Segmentation
Received date: 2021-01-26
Online published: 2021-08-26
前列腺区域的精确分割是提高计算机辅助前列腺癌诊断准确率的重要前提.本文提出了一种新的精确的前列腺区域分割模型，分为4个步骤：首先，读取T2加权磁共振（MR）图像；其次，利用半径为5个像素的8邻域模板（8x5）的局部二值模式（LBP）特征模板计算前列腺磁共振图像的LBP特征图；然后，利用改进的距离正则化水平集（DRLSE）模型对特征图进行分割，提取前列腺粗轮廓；最后将原始水平集能量函数进行优化，构造一个新的能量函数，提取局部灰度信息和梯度信息，并在此新的能量函数的基础上，将粗轮廓迭代演化为最终的细轮廓.本文将该模型在203组来自于国际光学与光子学学会-美国医学物理学家协会-国家癌症研究所（SPIE-AAPM-NCI）前列腺MR分类挑战数据库的T2W磁共振图像上进行了测试，并与医生手工分割结果进行了比较，结果表明本文提出模型得到的分割结果的Dice系数为0.94±0.01，相对体积差（RVD）为-1.21%±2.44%，95% Hausdorff距离（HD）为6.15±0.66 mm；与文献中现有的分割模型相比，使用本文提出的模型得到的前列腺区域分割结果更接近于手工分割的结果.
关键词： 局部灰度信息; 磁共振成像(MRI); 前列腺区域分割; 水平集; 计算机辅助诊断
闫士举 , 韩勇森 , 汤光宇 . 一种用于前列腺区域分割的改进水平集算法[J]. 波谱学杂志, 2021 , 38(3) : 356 -366 . DOI: 10.11938/cjmr20212885
Accurate segmentation of prostate region is an important prerequisite to improve the accuracy of computer-aided prostate cancer diagnosis. In this work, a new and accurate prostate segmentation algorithm is proposed and tested. The new algorithm consists of 4 steps: reading T2-weighted magnetic resonance images, calculating local binary pattern (LBP) feature map of prostate magnetic resonance images by using an 8x5 LBP feature template, segmenting the feature map with the improved distance regularization level set evolution (DRLSE) algorithm, and extracting coarse contour of the prostate. A new energy function is constructed to extract local gray scale information and gradient information, and the coarse contour is iteratively developed into the final fine prostate contour on the basis of this new energy function. The algorithm was tested with the SPIE-AAPM-NCI Prostate MR Classification Challenge Database. The segmentation results of the proposed algorithm were compared with that of manual segmentation by doctors. The results showed that the Dice coefficient obtained by using the proposed algorithm was 0.94±0.01, with a relative volume difference (RVD) of -1.21%±2.44% and a 95% Hausdorff distance (HD) of 6.15±0.66 mm. Compared with the existing segmentation algorithms, the segmentation results obtained with the algorithm proposed in this paper are closer to the manual segmentation results.
| 1 | WORLD HEALTH ORGANIZATION. Latest global cancer data: Cancer burden rises to 19.3 million new cases and 10.0 million cancer deaths in 2020 questions and answers (Q & A)[OL]. https://www.iarc.fr/faq/latest-global-cancer-data-2020-qa. |
| 2 | ANAS E M A , MOUSAVI P , ABOLMAESUMI P . A deep learning approach for real time prostate segmentation in freehand ultrasound guided biopsy[J]. Med Image Anal, 2018, 48, 107- 116. |
| 3 | WANG L J , SU X Y , LI Y , et al. Segmentation of right ventricle in cardiac cine MRI using COLLATE fusion-based multi-atlas[J]. Chinese J Magn Reson, 2018, 35 (4): 407- 416. |
| 3 | 王丽嘉, 苏新宇, 李亚, 等. 基于COLLATE融合多图谱的心脏电影MRI右心室分割[J]. 波谱学杂志, 2018, 35 (4): 407- 416. |
| 4 | WANG X L . Application of histogram analysis of dynamic enhanced MRI quantitative parameter in the diagnosis of prostate cancer[J]. Chinese Journal of CT and MRI, 2020, 18 (12): 110- 113. |
| 4 | 王晓蕾. 动态增强MRI定量参数直方图分析在诊断前列腺癌中的应用[J]. 中国CT和MRI杂志, 2020, 18 (12): 110- 113. |
| 5 | PALUMBO P , MANETTA R , IZZO A , et al. Biparametric (bp) and multiparametric (mp) magnetic resonance imaging (MRI) approach to prostate cancer disease: a narrative review of current debate on dynamic contrast enhancement[J]. Gland Surg, 2020, 9 (6): 2235- 2247. |
| 6 | LIU K W , LIU Z L , WANG X Y , et al. Prostate cancer diagnosis based on cascaded convolutional neural networks[J]. Chinese J Magn Reson, 2020, 37 (2): 152- 161. |
| 6 | 刘可文, 刘紫龙, 汪香玉, 等. 基于级联卷积神经网络的前列腺磁共振图像分类[J]. 波谱学杂志, 2020, 37 (2): 152- 161. |
| 7 | VAFAIE R, ALIREZAIE J, BABYN P. Fully automated model-based prostate boundary segmentation using markov random field in ultrasound images[C]//Fremantle, WA, Australia: International Conference on Digital Image Computing Techniques and Applications (DICTA), 2012. |
| 8 | KWAK J T , SANKINENI S , XU S , et al. Correlation of magnetic resonance imaging with digital histopathology in prostate[J]. Int J Comput Ass Rad, 2016, 11 (4): 657- 666. |
| 9 | QIAN C J , WANG L , GAO Y Z , et al. In vivo MRI based prostate cancer identification with random forests and auto-context model[J]. Comput Med Imag Grap, 2014, 52, 44- 57. |
| 10 | KORSAGER A S , FORTUNATI V , FEDDE VDL , et al. The use of atlas registration and graph cuts for prostate segmentation in magnetic resonance images[J]. Med Phys, 2015, 42 (4): 1614- 1624. |
| 11 | TIAN Z Q , LIU L Z , ZHANG Z F , et al. Superpixel-based segmentation for 3D prostate MR images[J]. IEEE Trans Med Imag, 2016, 35 (3): 791- 801. |
| 12 | LI C M , XU C Y , GUI C F , et al. Distance regularized level set evolution and its application to image segmentation[J]. IEEE T Image Process, 2010, 19 (12): 3243- 3254. |
| 13 | ZHANG Y D , PENG J C , LIU G , et al. Research on the segmentation method of prostate magnetic resonance image based on level set[J]. Chinese Journal of Scientific Instrument, 2017, 38 (2): 416- 424. |
| 13 | 张永德, 彭景春, 刘罡, 等. 基于水平集的前列腺磁共振图像分割方法研究[J]. 仪器仪表学报, 2017, 38 (2): 416- 424. |
| 14 | ZHU Z H , YAN S J , RUAN Y , et al. Segmentation of prostate magnetic resonance images based on an improved distance regularized level set evolution (DRLSE) model[J]. Chinese J Magn Reson, 2020, 37 (4): 447- 455. |
| 14 | 朱泽华, 闫士举, 阮渊, 等. 基于改进DRLSE模型的前列腺磁共振图像分割[J]. 波谱学杂志, 2020, 37 (4): 447- 455. |
| 15 | LI C M , KAO C Y , GORE J C , et al. Minimization of region-scalable fitting energy for image segmentation[J]. IEEE T Image Process, 2008, 17 (10): 1940- 1949. |
| 16 | JIANG H Y , FENG R J , GAO X H . Level set based on signed pressure force function and its application in liver image segmentation[J]. Wuhan University Journal of Natural Sciences, 2011, 16 (3): 265- 270. |
| 17 | ZHANG K H , SONG H H , ZHAND L . ZHANG. Active contours driven by local image fitting energy[J]. Pattern Recogn, 2010, 43 (4): 1199- 1206. |
| 18 | KARIMI D , ZENG Q , MATHUR P , et al. Accurate and robust deep learning-based segmentation of the prostate clinical target volume in ultrasound images[J]. Med Image Anal, 2019, 57, 186- 196. |
| 19 | MILLETARI F, NAVAB N, AHMADI S A. V-net: fully convolutional neural networks for volumetric medical image segmentation[C]//Stanford, CA, USA: 2016 Fourth International Conference on 3D Vision (3DV), 2016. |
| 20 | OJALA T , PIETIKAINEN M , MAENPAA T . Multiresolution gray-scale and rotation invariant texture classification with local binary patterns[J]. IEEE T Pattern Anal, 2002, 7 (24): 971- 987. |
| 21 | OSHER S , FEDKIW R . Level set methods and dynamic implicit surfaces[M]. New York: Springer-Verlag, 2002. |
| 22 | ZHAO H K , CHAN T , MERRIMAN B , et al. A variational level setapproach to multiphase motion[J]. J Comput Phys, 1996, 127 (1): 179- 195. |
| 23 | OTSU N . A threshold selection method from gray-level histogram[J]. IEEE Transactions on Systems, Man, and Cybernetics, 1979, 9 (1): 62- 66. |
| 24 | LI C M , HUANG R , DING Z H , et al. A level set method for image segmentation in the presence of intensity inhomogeneities with application to MRI[J]. IEEE T Image Process, 2011, 20 (7): 2007- 2016. |
| 25 | XU C Y, YEZZI A, PRINCE J L. On the relationship between parametric and geometric active contours[C]//Pacific Grove, CA, USA: Conference Record of the Thirty-Fourth Asilomar Conference on Signals, Systems and Computers, 2000. doi: 10.1109/ACSSC.2000.911003. |
| 26 | ZHANG K , ZHANG L , SONG H , et al. Active contours with selective local or global segmentation: A new formulation and level set method[J]. Image Vision Comput, 2010, 28 (4): 668- 676. |
| 27 | CHAN T F , VESE L A . Active contours without edges[J]. IEEE T Image Process, 2001, 10 (2): 266- 277. |
| 28 | GEERT L , ROBERT T , WENDY V D V , et al. Evaluation of prostate segmentation algorithms for MRI: The PROMISE12 challenge[J]. Med Image Anal, 2014, 18 (2): 359- 373. |
| 29 | MAHAPATRA D , BUHMANN J M . Visual saliency-based active learning for prostate magnetic resonance imaging segmentation[J]. J Med Imaging, 2016, 3 (1): 014003. |
| 30 | KARIMI D , SAMEI G , KESCH C , et al. Prostate segmentation in MRI using a convolutional neural network architecture and training strategy based on statistical shape models[J]. Int J Comput Ass Rad, 2018, 13 (4): 1- 9. |
/
| 〈 |
|
〉 |