Placental Maturity Evaluation via Feature Fusionand Discriminative Learning
Li Wanjun1, Wang Tianfu1, Ni Dong1, Chen Siping, Lei Baiying1*, YaoYuan2*
1(Department of Biomedical Engineering, School of Medicine, Shenzhen University, National-Regional Key TechnologyEngineering Laboratory for Medical Ultrasound, Guangdong Key Laboratory for Biomedical Measurements and Ultrasound Imaging, Shenzhen 518060, Guangdong, China) 2(Department of Ultrasound, Affiliated Shenzhen Maternal and Child Healthcare, Hospital of Nanfang Medical University, Shenzhen518060, Guangdong, China)
Abstract：The error of placental maturity classification may lead to the occurrence of small gestational age (SGA), stillbirth, dead fetus, etc. Currently, placental maturity evaluation mainly depends on the clinician's experience and observation. In this paper, we proposed a novel method to evaluate the placental maturity automatically by feature fusion and discriminative learning. Specifically, we extracted both the gray-scale intensity and blood flow information by the visual feature detector and descriptor from a total of 544 B-mode gray-scale ultrasound (US) images and color doppler energy (CDE) images. After fusing information, we applied the feature encoding method to improve the staging performance using discriminative learning technique. Comparing the test results with the result of the clinicians, we obtained a result with the accuracy of 92.7%, the sensitivity of 91.1%, a specificity of 97.6%, and a mean average precision of 97.3%. The experimental results showed that the proposed method achieved promising performance for placental maturity automatic classification.
[1] 张兰珍,谭丽君,李大慈.胎盘成熟度分级与胎盘功能的相关性探讨 [J]. 中山医科大学学报, 2001, 22(5): 386-389. [2] 李秋萍,陈汝芳,谢咏. 超声检查胎盘成熟度预测胎盘功能的临床价值[J].中国现代医学杂志, 2001, 11(5): 78-79. [3] Grannum PA, Berkowitz RL, Hobbins JC. The ultrasonic changes in the maturing placenta and their relation to fetal pulmonic maturity[J]. American Journal of Obstetrics & Gynecology, 1979, 133(8): 915-922. [4] Linares PA, McCullagh PJ, Black ND, et al. Feature selection for the characterization of ultrasonic images of the placenta using texture classification[C] //2004 IEEE International Symposium on Biomedical Imaging: Nano to Macro. Arlington: IEEE, 2004: 1147-1150. [5] Liu Z, Zheng H, Lin S. Application of multi-classification support vector machine in the B-placenta image classification[C]// 2009 International Conference on Computational Intelligence and Software Engineering. Wuhan: IEEE, 2009: 1-4. [6] Mao J, Yang X, Tian X, et al. The Study of automatic classification for ultrasound placenta images based on adaptive multiple neural networks[J]. Journal of Image and Graphics, 2006, 11(7): 965-970. [7] Lei B, Li X, Yao Y, et al. Automatic grading of placental maturity based on LIOP and fisher vector[C]// 2014 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society. Chicago: IEEE, 2014: 4671-4674. [8] Li X, Yao Y, Ni D, et al. Automatic staging of placental maturity based on dense descriptor[J]. Biomedical Materials and Engineering, 2014, 24(6): 2821-2829. [9] Lei B, Yao Y, Chen S, et al. Discriminative learning for automatic staging of placental maturity via multi-layer Fisher vector[J]. Scientific Reports, 2015, 5(12818):1-11. [10] Burton GJ, Charnock-Jones DS, Jauniaux E. Regulation of vascular growth and function in the human placenta[J]. Reproduction, 2009, 138(6): 895-902. [11] Guiot C, Gaglioti P, Oberto M, et al. Is three-dimensional power Doppler ultrasound useful in the assessment of placental perfusion in normal and growth-restricted pregnancies?[J]. Ultrasound in Obstetrics & Gynecology, 2008, 31(2): 171-176. [12] Bude RO, Rubin JM. Power Doppler sonography[J]. Radiology, 1996, 200(1): 21-23. [13] Lei B, Tan EL, Chen S, et al. Saliency-driven image classification method based on histogram mining and image score[J]. Pattern Recognition, 2015, 48(8): 2567-2580. [14] Lazebnik S, Schmid C, Ponce J. Beyond bags of features: Spatial pyramid matching for recognizing natural scene categories[C]//2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.New York: IEEE, 2006: 2169-2178. [15] Jégou H, Perronnin F, Douze M, et al. Aggregating local image descriptors into compact codes[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2012, 34(9): 1704-1716. [16] Sánchez J, Perronnin F, Mensink T, et al. Image classification with the fisher vector: Theory and practice[J]. International Journal of Computer Vision, 2013, 105(3): 222-245. [17] Lei B, Tan EL, Chen S, et al. Automatic recognition of fetal facial standard plane in ultrasound image via Fisher vector[J]. PLoS ONE, 2015, 10(5): e0121838. [18] Shen H, Tao D, Ma D. Multiview locally linear embedding for effective medical image retrieval[J]. PLoS ONE, 2013, 8(12): e82409. [19] Chu C, Hsu AL, Chou KH, et al. Does feature selection improve classification accuracy? Impact of sample size and feature selection on classification using anatomical magnetic resonance images[J]. Neuroimage, 2012, 60(1): 59-70. [20] Perronnin F, Sánchez J, Mensink T. Lecture notes in computer science: improving the fisher kernel for large-scale image classification [M]//Computer Vision-ECCV 2010. Beilin: Springer Berlin Heidelberg, 2010: 143-156.
