Retinal Vessel Segmentation Method with Efficient Hybrid Features Fusion
Funds:
The National Natural Science Foundation of China (61201360)
-
摘要: 將機(jī)器學(xué)習(xí)運(yùn)用到視網(wǎng)膜血管分割當(dāng)中已成為一種趨勢(shì)，然而選取什么特征作為血管與非血管的特征仍為眾所思考的問(wèn)題。該文利用將血管像素與非血管像素看作二分類(lèi)的原理，提出一種混合的5D特征作為血管像素與非血管像素的表達(dá)，從而能夠簡(jiǎn)單快速地將視網(wǎng)膜血管從背景中分割開(kāi)來(lái)。其中5D特征向量包括CLAHE (Contrast Limited Adaptive Histgram Equalization)，高斯匹配濾波，Hesse矩陣變換，形態(tài)學(xué)底帽變換，B-COSFIRE(Bar-selective Combination Of Shifted FIlter REsponses)，通過(guò)將融合特征輸入SVM(支持向量機(jī))分類(lèi)器訓(xùn)練得到所需的模型。通過(guò)在DRIVE和STARE數(shù)據(jù)庫(kù)進(jìn)行實(shí)驗(yàn)分析，利用Se, Sp, Acc, Ppv, Npv, F1-measure等常規(guī)評(píng)價(jià)指標(biāo)來(lái)檢測(cè)分割效果，其中平均準(zhǔn)確率分別達(dá)到0.9573和0.9575，結(jié)果顯示該融合方法比單獨(dú)使用B-COSFIRE或者其他目前所提出的融合特征方法更準(zhǔn)確有效。
-
關(guān)鍵詞:
- 機(jī)器學(xué)習(xí) /
- 視網(wǎng)膜 /
- 血管分割 /
- 特征向量 /
- 支持向量機(jī)
Abstract: How to apply machine learning to retinal vessel segmentation effectively has become a trend, however, choosing what kind of features for the blood vessels is still a problem. In this paper, the blood vessels of pixels are regarded as a theory of binary classification, and a hybrid 5D features for each pixel is put forward to extract retinal blood vessels from the background simplely and quickly. The 5D eigenvector includes Contrast Limited Adaptive Histgram Equalization (CLAHE), Gaussian matched filter, Hessian matrix transform, morphological bottom hat transform and Bar-selective Combination Of Shifted Filter Responses (B-COSFIRE). Then the fusion features are input into the Support Vector Machine (SVM) classifier to train a model that is needed. The proposed method is evaluated on two publicly available datasets of DRIVE and STARE, respectively. Se, Sp, Acc, Ppv, Npv, F1-measure are used to test the proposed method, and average classification accuracies are 0.9573 and 0.9575 on the DRIVE and STARE datasets, respectively. Performance results show that the fusion method also outperform the state-of-the-art method including B-COSFIRE and other currently proposed fusion features method.-
Key words:
- Machine learning /
- Retina /
- Vessel segmentation /
- Feature vetor /
- Support Vector Machine (SVM)
-
揚(yáng)琴, 趙景秀. 基于主成分分析的血管分割算法[J]. 電子技術(shù)設(shè)計(jì)與應(yīng)用, 2016, 3: 66-70. YANG Qin and ZHAO Jingxiu. Blood vessel segmentation algorithm based on principal components analysis[J]. Electronics Design Application, 2016, 45(3): 66-70. SINGH N P and SRIVASTAVA R. Retinal blood vessels segmentation by using Gumbel probability distributiaon function based matched filter[J]. Comput Methods Progr Biomed, 2016, 129(C): 40-50. doi: 10.1016/j.cmpb.2016.03. 001. FRAZ M M, BASIT A, and BARMAN S A. Application of morphological bit planes in retinal blood vessel extration[J]. Digit Imaging. 2013, 26(2): 274-286. doi: 10.1007/s10278-012- 9513-3. FRAZ M M, REMAGNINO P, HOPPE A, et al. Blood vessel segmentation methodologies in retinalimages-A survey[J]. Computer Methods and Programs Biomedicine, 2012, 108(1): 407-433. doi: 10.1016/j.cmpb.2012.03009. 吳奎, 蔡冬梅, 賈鵬, 等. 基于2D Gabor小波與組合線檢測(cè)算子的視網(wǎng)膜血管分割[J]. 科學(xué)技術(shù)與工程, 2016, 16(12): 106-112. WU Kui, CAI Dongmei, JIA Peng, et al. Retinal vessel segmentation based on 2D Gabor wavelet and combined line operators[J]. Science Technology and Engineering, 2016, 16(12): 106-112. 于輝, 王小鵬. 基于HESSIAN增強(qiáng)和形態(tài)學(xué)尺度空間的視網(wǎng)膜血管分割[J]. 計(jì)算機(jī)應(yīng)用與軟件, 2016, 33(8): 200-205. YU Hui and WANG Xiaopeng. Retinal vessels segmentation based on Hessian enhancement and morphological scale space[J]. Computer Applications and Software, 2016, 33(8): 200-205. WAHEED Z, AKRAM M U, WAHEED A, et al. Person identification using vascular and non-vascular retinal feature[J]. Computers and Electrical Engineering, 2016, 53: 359-371. doi: 10.1016/j.compeleceng.2016.03.010. RICCI E and PERFETTI R. Retinal blood vessel segmentation using line operators and support vector classification[J]. IEEE Transactions on Medical Imaging, 2007, 26(10): 1357-1365. doi: 10.1109/TMI.2007.8985551. FRAZ M M, REMAGNINO P, HOPPE A, et al. An ensemble classification-based approach applied to retinal blood vessel segmentation[J]. IEEE Transactions on Biomedical Engineering, 2012, 59(9): 2538-2548. doi: 10.1109/TBME. 2012.2205687. AZZOPARDI G, STRISCIUGLIO N, VENTO M, et al. Trainable COSFIRE filters for vessel delineation with application to retinal images[J]. Medical Image Analysis, 2015, 19(1): 46-57. doi: 10.1016/j.media.2014.08.002. STRISCIUGLIO N, AZZOPARDI G, VENTO M, et al. Supervised vessel delineation in retinal fundus images with the automatic selection of B-COSFIRE filters[J]. Machine Vision and Applications, 2016, 27(8): 1137-1149. doi: 10.1007 //s00138-016-0781-7. ASLANI S and SARNEL H. A new supervised retinal vessel segmentation method based on robust hybrid feature[J]. Biomedical Singal Processing and Control, 2016, 30: 1-12. doi: 10.1016/j.bspc.2016.05006. -
計(jì)量
- 文章訪問(wèn)數(shù): 1382
- HTML全文瀏覽量: 229
- PDF下載量: 351
- 被引次數(shù): 0
下載:
