Abstract:Automatic segmentation and classification of lesions are two clinically significant tasks in the computer-aided diagnosis of skin diseases. Both tasks are challenging due to the nonnegligible lesion differences in dermoscopic images from different patients. In this paper, we propose a novel pipeline to efficiently implement skin lesions' segmentation and classification tasks, which consists of a segmentation network and a classification network. To improve the performance of the segmentation network, we propose a novel module of Multi -Scale Holistic Feature Exploration (MSH) to thoroughly exploit perceptual clues latent among multi-scale feature maps as synthesized by the decoder. The MSH module enables holistic exploration of features across multiple scales to more effectively support downstream image analysis tasks. To boost the performance of the classification network, we propose a novel module of Cross-Modality Collaborative Feature Exploration (CMC) to discover latent discriminative features by collaboratively exploiting potential relationships between cross -modal features of dermoscopic images and clinical metadata. The CMC module enables dynamically capturing versatile interaction effects among cross-modal features during the model's representation learning procedure by discriminatively and adaptively learning the interaction weight associated with each crossmodality feature pair. In addition, to effectively reduce background noise and boost the lesion discrimination ability of the classification network, we crop the images based on lesion masks generated by the best segmentation model. We evaluate the proposed pipeline on the four public skin lesion datasets, where the ISIC 2018 and PH2 are for segmentation, and the ISIC 2019 and ISIC 2020 are combined into a new dataset, ISIC 2019&2020, for classification. It achieves a Jaccard index of 83.31% and 90.14% in skin lesion segmentation, an AUC of 97.98% and an Accuracy of 92.63% in skin lesion classification, which is superior to the performance of representative state-of-the-art skin lesion segmentation and classification methods. Last but not least, the new model for segmentation utilizes much fewer model parameters (3.3 M) than its peer approaches, leading to a greatly reduced number of labeled samples required for model training, which obtains substantially stronger robustness than its peers.

A Two-Stage Method for Skin Lesion Analysis

A Detection and Segmentation Architecture for Skin Lesion Segmentation on Dermoscopy Images

Attention-Guided Network with Densely Connected Convolution for Skin Lesion Segmentation

Skin Lesion Analysis Towards Melanoma Detection Using Deep Learning Network

ISIC 2017 - Skin Lesion Analysis Towards Melanoma Detection

Deep Learning-Based Semantic Segmentation of Non-Melanocytic Skin Tumors in Whole-Slide Histopathological Images.

A Novel Multi-task Deep Learning Model for Skin Lesion Segmentation and Classification

Unsupervised sub-segmentation for pigmented skin lesions.

Skin Lesion Analysis Towards Melanoma Detection via End-to-end Deep Learning of Convolutional Neural Networks

Learning from Dermoscopic Images in Association with Clinical Metadata for Skin Lesion Segmentation and Classification

Using Deep Learning Method for Classification: A Proposed Algorithm for the ISIC 2017 Skin Lesion Classification Challenge

Dermoscopic Image Analysis for ISIC Challenge 2018.

ISIC 2017 Skin Lesion Segmentation Using Deep Encoder-Decoder Network

Automatic skin lesion segmentation with fully convolutional-deconvolutional networks

Deep Attention-Guided Fusion Network for Lesion Segmentation.

Automated lesion segmentation and dermoscopic feature segmentation for skin cancer analysis

Portable Skin Lesion Segmentation System with Accurate Lesion Localization Based on Weakly Supervised Learning

Comparative Analysis of Automatic Skin Lesion Segmentation with Two Different Implementations

LesionSeg: Semantic segmentation of skin lesions using Deep Convolutional Neural Network

SEACU-Net: Attentive ConvLSTM U-Net with squeeze-and-excitation layer for skin lesion segmentation

Complementary Network with Adaptive Receptive Fields for Melanoma Segmentation