Complementary Network with Adaptive Receptive Fields for Melanoma Segmentation

TL;DR

This paper introduces a novel complementary network with adaptive receptive fields and mutual learning for improved melanoma segmentation in dermoscopic images, addressing hole and shrink problems to enhance boundary accuracy and overall performance.

Contribution

It proposes a dual-network architecture with adaptive atrous convolution and a knowledge aggregation module, along with a mutual loss for semi-supervised learning, advancing melanoma segmentation techniques.

Findings

Achieved a dice coefficient of 86.4% on ISIC 2018 dataset.

Outperformed state-of-the-art melanoma segmentation methods.

Enhanced boundary sensitivity and segmentation robustness.

Abstract

Automatic melanoma segmentation in dermoscopic images is essential in computer-aided diagnosis of skin cancer. Existing methods may suffer from the hole and shrink problems with limited segmentation performance. To tackle these issues, we propose a novel complementary network with adaptive receptive filed learning. Instead of regarding the segmentation task independently, we introduce a foreground network to detect melanoma lesions and a background network to mask non-melanoma regions. Moreover, we propose adaptive atrous convolution (AAC) and knowledge aggregation module (KAM) to fill holes and alleviate the shrink problems. AAC explicitly controls the receptive field at multiple scales and KAM convolves shallow feature maps by dilated convolutions with adaptive receptive fields, which are adjusted according to deep feature maps. In addition, a novel mutual loss is proposed to utilize the dependency between the foreground and background networks, thereby enabling the reciprocally influence within these two networks. Consequently, this mutual training strategy enables the semi-supervised learning and improve the boundary-sensitivity. Training with Skin Imaging Collaboration (ISIC) 2018 skin lesion segmentation dataset, our method achieves a dice co-efficient of 86.4% and shows better performance compared with state-of-the-art melanoma segmentation methods.