Cascaded atrous convolution and spatial pyramid pooling for more accurate tumor target segmentation for rectal cancer radiotherapy

Convolutional neural networks (CNNs) have become the state-of-the-art method for medical segmentation. However, repeated pooling and striding operations reduce the feature resolution, causing loss of detailed information. Additionally, tumors of different patients are of different sizes. Thus, small tumors may be ignored while big tumors may exceed the receptive fields of convolutions. The purpose of this study is to further improve the segmentation accuracy using a novel CNN (named CAC–SPP) with cascaded atrous convolution (CAC) and a spatial pyramid pooling (SPP) module. This work is the first attempt at applying SPP for segmentation in radiotherapy. We improved the network based on ResNet-101 yielding accuracy gains from a greatly increased depth. We added CAC to extract a high-resolution feature map while maintaining large receptive fields. We also adopted a parallel SPP module with different atrous rates to capture the multi-scale features. The performance was compared with the widely adopted U-Net and ResNet-101 with independent segmentation of rectal tumors for two image sets, separately: (1) 70 T2-weighted MR images and (2) 100 planning CT images. The results show that the proposed CAC–SPP outperformed the U-Net and ResNet-101 for both image sets. The Dice similarity coefficient values of CAC–SPP were 0.78 ± 0.08 and 0.85 ± 0.03, respectively, which were higher than those of U-Net (0.70 ± 0.11 and 0.82 ± 0.04) and ResNet-101 (0.76 ± 0.10 and 0.84 ± 0.03). The segmentation speed of CAC–SPP was comparable with ResNet-101, but about 36% faster than U-Net. In conclusion, the proposed CAC–SPP, which could extract high-resolution features with large receptive fields and capture multi-scale context yields, improves the accuracy of segmentation performance for rectal tumors.