Retinal Microaneurysm Detection Using Clinical Report Guided Multi-sieving CNN
Timely detection and treatment of microaneurysms (MA) is a critical step to prevent the development of vision-threatening eye diseases such as diabetic retinopathy. However, detecting MAs in fundus images is a highly challenging task due to the large variation of imaging conditions. In this paper, we focus on developing an interleaved deep mining technique to cope intelligently with the unbalanced MA detection problem. Specifically, we present a clinical report guided multi-sieving convolutional neural network (MS-CNN) which leverages a small amount of supervised information in clinical reports to identify the potential MA regions via a text-to-image mapping in the feature space. These potential MA regions are then interleaved with the fundus image information for multi-sieving deep mining in a highly unbalanced classification problem. Critically, the clinical reports are employed to bridge the semantic gap between low-level image features and high-level diagnostic information. Extensive evaluations show our framework achieves 99.7% precision and 87.8% recall, comparing favorably with the state-of-the-art algorithms. Integration of expert domain knowledge and image information demonstrates the feasibility to reduce the training difficulty of the classifiers under extremely unbalanced data distribution.
This work is supported by National High-tech R&D Program of China (863 Program) (2015AA015904), NSFC (61572316, 61671290, 61525106), National Key R&D Program of China (2016YFC1300302), Key Program for International S&T Cooperation Project (2016YFE0129500) of China, Science and Technology Commission of Shanghai Municipality (16DZ0501100), and Interdisciplinary Program of Shanghai Jiao Tong University (14JCY10).
- 2.UK Prospective Diabetes Study Group. Tight blood pressure control and risk of macrovascular and microvascular complications in type 2 diabetes: Ukpds 38. BMJ: British Medical Journal, pp. 703–713 (1998)Google Scholar
- 4.Hofmanninger, J., Langs, G.: Mapping visual features to semantic profiles for retrieval in medical imaging. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 457–465 (2015)Google Scholar
- 5.Achanta, R., Shaji, A., Smith, K., Lucchi, A., Fua, P., Susstrunk, S.: SLIC Superpixels. EPFL Technical report 149300, p. 15, June 2010Google Scholar
- 6.Krizhevsky, A., Sulskever, I., Hinton, G.E.: ImageNet classification with deep convolutional neural networks. In: Advances in Neural Information and Processing Systems (NIPS), pp. 1–9 (2012)Google Scholar
- 7.Ozuysal, M., Calonder, M., Lepetit, V., Fua, P.: Fast keypoint recognition using random ferns. IEEE Trans. Pattern Anal. Mach. Intell. 32(Xx), 1–14 (2010)Google Scholar
- 9.Kauppi, T., et al.: The DIARETDB1 diabetic retinopathy database and evaluation protocol. In: BMVC, pp. 1–10 (2007)Google Scholar
- 11.Mizutani, A., Muramatsu, C., Hatanaka, Y., Suemori, S., Hara, T., Fujita, H.: Automated microaneurysm detection method based on double ring filter in retinal fundus images. 7260:72601N–72601N-8 (2009)Google Scholar
- 12.Aravind, C., Ponnibala, M., Vijayachitra, S.: Automatic detection of microaneurysms and classification of diabetic retinopathy images using SVM technique. In: IJCA Proceedings on International Conference on Innovations in Intelligent Instrumentation, Optimization and Electrical Sciences ICIIIOES 11, pp. 18–22 (2013)Google Scholar
- 14.Sehirli, E., Turan, M.K., Dietzel, A.: Automatic detection of microaneurysms in RGB retinal fundus images. Studies 1(8) (2015)Google Scholar