RFC-Net: Learning High Resolution Global Features for Medical Image Segmentation on a Computational Budget
This work addresses computational efficiency in medical image segmentation for polyp detection, which is incremental as it builds on existing CNN architectures.
The authors tackled the problem of information loss in medical image segmentation by proposing RFC-Net, which uses a loose dense connection strategy and m-way tree structure to learn high-resolution global features with reduced computational cost, achieving state-of-the-art performance on Kvasir and CVC-ClinicDB benchmarks for polyp segmentation.
Learning High-Resolution representations is essential for semantic segmentation. Convolutional neural network (CNN)architectures with downstream and upstream propagation flow are popular for segmentation in medical diagnosis. However, due to performing spatial downsampling and upsampling in multiple stages, information loss is inexorable. On the contrary, connecting layers densely on high spatial resolution is computationally expensive. In this work, we devise a Loose Dense Connection Strategy to connect neurons in subsequent layers with reduced parameters. On top of that, using a m-way Tree structure for feature propagation we propose Receptive Field Chain Network (RFC-Net) that learns high resolution global features on a compressed computational space. Our experiments demonstrates that RFC-Net achieves state-of-the-art performance on Kvasir and CVC-ClinicDB benchmarks for Polyp segmentation.