Accurate Retinal Vessel Segmentation via Octave Convolution Neural Network

Retinal vessel segmentation is a crucial step in diagnosing and screening various diseases, including diabetes, ophthalmologic diseases, and cardiovascular diseases. In this paper, we propose an effective and efficient method for vessel segmentation in color fundus images using encoder-decoder based octave convolution network. Compared with other convolution networks utilizing vanilla convolution for feature extraction, the proposed method adopts octave convolution for learning multiple-spatial-frequency features, thus can better capture retinal vasculatures with varying sizes and shapes. It is empirically demonstrated that the feature maps of low-frequency kernels respond mainly on the major vascular tree, whereas the high-frequency feature maps can better capture the fine details of thin vessels. To provide the network the capability of learning how to decode multifrequency features, we extend octave convolution and propose a novel operation named octave transposed convolution with a similar multifrequency approach as the octave convolution. We also propose a novel architecture of fully convolutional neural network named Octave UNet based on the encoder-decoder architecture of UNet. The proposed Octave UNet can generate high-resolution vessel segmentation in single forward feeding. The proposed method is evaluated on four publicly available datasets, including DRIVE, STARE, CHASE_DB1, and HRF. Extensive experimental results demonstrate that the proposed approach achieves better or comparable performance to the state-of-the-art methods with fast processing speed.