Abstract:Skin cancer images have hair occlusion problems, which greatly affects the accuracy of diagnosis and classification. Current dermoscopic hair removal methods use segmentation networks to locate hairs, and then uses repair networks to perform image repair. However, it is difficult to segment hair and capture the overall structure between hairs because of the hair being thin, unclear, and similar in color to the entire image. When conducting image restoration tasks, the only available images are those obstructed by hair, and there is no corresponding ground truth (supervised data) of the same scene without hair obstruction. In addition, the texture information and structural information used in existing repair methods are often insufficient, which leads to poor results in skin cancer image repair. To address these challenges, we propose the intersection-union dual-stream cross-attention Lova-SwinUnet (IUDC-LS). Firstly, we propose the Lova-SwinUnet module, which embeds Lovasz loss function into Swin-Unet, enabling the network to better capture features of various scales, thus obtaining better hair mask segmentation results. Secondly, we design the intersection-union (IU) module, which takes the mask results obtained in the previous step for pairwise intersection or union, and then overlays the results on the skin cancer image without hair to generate the labeled training data. This turns the unsupervised image repair task into the supervised one. Finally, we propose the dual-stream cross-attention (DC) module, which makes texture information and structure information interact with each other, and then uses cross-attention to make the network pay attention to the more important texture information and structure information in the fusion process of texture information and structure information, so as to improve the effect of image repair. The experimental results show that the PSNR index and SSIM index of the proposed method are increased by 5.4875 and 0.0401 compared with the other common methods. Experimental results unequivocally demonstrate the effectiveness of our approach, which serves as a potent tool for skin cancer detection, significantly surpassing the performance of comparable methods.

An Approach to Unsupervised Hair Removal from Skin Melanoma Image

Oxygenation Using Tidal Volume Breathing After Maximal Exhalation

Research on hair removal algorithm of dermatoscopic images based on maximum variance fuzzy clustering and optimization Criminisi algorithm

Unsupervised sub-segmentation for pigmented skin lesions.

Distribution Quantification on Dermoscopy Images for Computer-Assisted Diagnosis of Cutaneous Melanomas

Intersection-union dual-stream cross-attention Lova-SwinUnet for skin cancer hair segmentation and image repair

Towards Better Dermoscopic Image Feature Representation Learning for Melanoma Classification

Automated lesion segmentation and dermoscopic feature segmentation for skin cancer analysis

Recovering Skin Reflectance And Geometry For Diagnosis Of Melanoma

An Encoder-Decoder CNN for Hair Removal in Dermoscopic Images

Two-Stage Deep Neural Network Via Ensemble Learning for Melanoma Classification

The effect of hair removal and filtering on melanoma detection: a comparative deep learning study with AlexNet CNN

Automatic Melanoma Detection using Discrete Cosine Transform Features and Metadata on Dermoscopic Images

Complementary Network with Adaptive Receptive Fields for Melanoma Segmentation

An efficient multi-level pre-processing algorithm for the enhancement of dermoscopy images in melanoma detection

Psychometric assessment of the Health Care Alliance Questionnaire with women in prenatal care

No-reference hair occlusion assessment for dermoscopy images based on distribution feature

Automatic Detection and Quantitative Evaluation of Skin Hair Follicles Based on Mueller Matrix Polarimetry

Skin Lesion Extraction Using Multiscale Morphological Local Variance Reconstruction Based Watershed Transform and Fast Fuzzy C-Means Clustering

Deep Hybrid Convolutional Neural Network for Segmentation of Melanoma Skin Lesion

Melanoma Detection Using XGB Classifier Combined with Feature Extraction and K-means SMOTE Techniques