This is the accompanying table for Mirikharaji et al., "A Survey on Deep Learning for Skin Lesion Segmentation", 2023. A downloadable version is also available as a Google Sheets document.
| Index | Paper Title | Year | Venue | Datasets | Architectural Modules | Loss Function | Performance (Jaccard) | Cross-data Evaluation | Augmentation | Post-processing | Code Provided |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 16 | Vesal et al., A multi-task framework for skin lesion detection and segmentation | 2018 | peer-reviewed conference | ISIC2017, PH2 | dilated conv., dense con., skip con. | Dice | 88.00% | ✓ | - | ✗ | ✗ |
| 36 | Goyal et al., Skin Lesion Segmentation in Dermoscopic Images with Ensemble Deep Learning Methods | 2019 | peer-reviewed journal | ISIC2017, PH2 | dilated conv., parallel m.s. conv., separable conv. | - | 79.34% | ✓ | - | ✓ | ✗ |
| 2 | He et al., Skin lesion segmentation via deep RefineNet | 2017 | peer-reviewed conference | ISIC2016, ISIC2017 | residual con., skip con., image pyramid | Dice, CE, DS | 75.80% | ✗ | rotation | ✓ | ✗ |
| 17 | Venkatesh et al., A deep residual architecture for skin lesion segmentation | 2018 | peer-reviewed conference | ISIC2017 | residual con., skip con. | Jaccard | 76.40% | ✗ | rotation, flipping, translation, scaling | ✓ | ✗ |
| 37 | Azad et al., Bi-Directional ConvLSTM U-Net with Densley Connected Convolutions | 2019 | peer-reviewed conference | ISIC2018 | skip con., dense con., recurrent CNN | CE | 74.00% | ✗ | - | ✗ | ✓ |
| 38 | Alom et al., Recurrent residual U-Net for medical image segmentation | 2019 | peer-reviewed journal | ISIC2017 | skip con., residual con., recurrent CNN | CE | 75.68% | ✗ | - | ✗ | ✗ |
| 39 | Yuan and Lo, Improving Dermoscopic Image Segmentation with Enhanced Convolutional-Deconvolutional Networks | 2019 | peer-reviewed journal | ISIC2017 | - | Tanimoto | 76.50% | ✗ | rotation, flipping, shifting, scaling, random normaliz. | ✓ | ✗ |
| 40 | Goyal et al., Skin lesion boundary segmentation with fully automated deep extreme cut methods | 2019 | peer-reviewed conference | ISIC2017, PH2 | dilated conv., parallel m.s. conv. | WCE | 82.20% | ✓ | - | ✗ | ✗ |
| 18 | Yang et al., Skin lesion analysis by multi-target deep neural networks | 2018 | non peer-reviewed technical report | ISIC2017 | skip con., parallel m.s. conv. | - | 74.10% | ✗ | rotation, flipping | ✗ | ✗ |
| 19 | Sarker et al., SLSDeep: Skin lesion segmentation based on dilated residual and pyramid pooling networks | 2018 | peer-reviewed conference | ISIC2016, ISIC2017 | skip con., residual con., dilated conv., pyramid pooling | CE, EPE | 78.20% | ✗ | rotation, scaling | ✗ | ✓ |
| 20 | Al-Masni et al., Skin lesion segmentation in dermoscopy images via deep full resolution convolutional networks | 2018 | peer-reviewed journal | ISIC2017, PH2 | - | CE | 77.10% | ✓ | rotation | ✗ | ✗ |
| 21 | Li et al., Deeply supervised rotation equivariant network for lesion segmentation in dermoscopy images | 2018 | peer-reviewed conference | ISIC2017 | skip con., residual con. | DS | 77.23% | ✗ | flipping, rotation | ✗ | ✓ |
| 41 | Bi et al., Step-wise integration of deep class-specific learning for dermoscopic image segmentation | 2019 | peer-reviewed journal | ISIC2016, ISIC2017, PH2 | skip con., residual con., | CE | 77.73% | ✓ | flipping, cropping | ✓ | ✗ |
| 3 | Bozorgtabar et al., Skin lesion segmentation using deep convolution networks guided by local unsupervised learning | 2017 | peer-reviewed journal | ISIC2016 | - | - | 80.60% | ✗ | rotation | ✗ | ✗ |
| 42 | Tschandl et al., Domain-specific classification-pretrained fully convolutional network encoders for skin lesion segmentation | 2019 | peer-reviewed journal | ISIC2017 | skip con. | CE, Jaccard | 76.80% | ✗ | flipping, rotation | ✓ | ✗ |
| 22 | Zeng and Zheng, Multi-scale fully convolutional DenseNets for automated skin lesion segmentation in dermoscopy images | 2018 | peer-reviewed conference | ISIC2017 | dense con., skip con., image pyramid | CE, l2, DS | 78.50% | ✗ | flipping, rotation | ✓ | ✗ |
| 43 | Li et al., Transformation-consistent self-ensembling model for semi-supervised medical image segmentation | 2021 | peer-reviewed journal | ISIC2017 | skip con., dense con., semi-supervised, ensemble | CE, l1 | 79.80% | ✗ | flipping, rotating, scaling | ✓ | ✗ |
| 23 | DeVries and Taylor, Leveraging uncertainty estimates for predicting segmentation quality | 2018 | non peer-reviewed technical report | ISIC2017 | skip con. | CE | 73.00% | ✗ | flipping, rotation | ✗ | ✗ |
| 44 | Zhang et al., Automatic skin lesion segmentation by coupling deep fully convolutional networks and shallow network with textons | 2019 | peer-reviewed journal | ISIC2016, ISIC2017 | skip con. | CE | 72.94% | ✗ | - | ✗ | ✗ |
| 45 | Baghersalimi et al., DermoNet: densely linked convolutional neural network for efficient skin lesion segmentation | 2019 | peer-reviewed journal | ISIC2016, ISIC2017, PH2 | skip con., residual con., dense con. | Tanimoto | 78.30% | ✓ | flipping, cropping | ✗ | ✗ |
| 78 | Hasan et al., DSNet: Automatic dermoscopic skin lesion segmentation | 2020 | peer-reviewed journal | ISIC2017, PH2 | skip con., dense con., separable conv. | CE, Jaccard | 77.50% | ✓ | rotation, zooming, shifting, flipping | ✗ | ✓ |
| 24 | Izadi et al., Generative adversarial networks to segment skin lesions | 2018 | peer-reviewed conference | DermoFit | skip con. | CE, ADV | 81.20% | ✗ | flipping, rotation, elastic deformation | ✗ | ✓ |
| 46 | Jiang et al., Decision-Augmented Generative Adversarial Network for Skin Lesion Segmentation | 2019 | peer-reviewed conference | ISIC2017 | residual con., dilated conv., GAN | ADV, l2 | 76.90% | ✗ | rotation, flipping | ✗ | ✗ |
| 47 | Tang et al., A multi-stage framework with context information fusion structure for skin lesion segmentation | 2019 | peer-reviewed conference | ISIC2016 | skip con. | Tanimoto, DS | 85.34% | ✗ | rotation, flipping | ✗ | ✗ |
| 48 | Bi et al., Improving Skin Lesion Segmentation via Stacked Adversarial Learning | 2019 | peer-reviewed conference | ISIC2017 | residual con. | CE | 77.14% | ✗ | GAN | ✗ | ✗ |
| 25 | Li et al., Dense deconvolutional network for skin lesion segmentation | 2018 | peer-reviewed journal | ISIC2016, ISIC2017 | skip con., residual con., dense con. | Jaccard, DS | 76.50% | ✗ | - | ✗ | ✗ |
| 26 | Mirikharaji and Hamarneh, Star shape prior in fully convolutional networks for skin lesion segmentation | 2018 | peer-reviewed conference | ISIC2017 | residual con. | CE, Star shape | 77.30% | ✗ | - | ✗ | ✗ |
| 49 | Abraham and Khan, A novel focal tversky loss function with improved attention U-Net for lesion segmentation | 2019 | peer-reviewed conference | ISIC2018 | skip con., image pyramid, attention | TV, Focal | 74.80% | ✗ | - | ✗ | ✓ |
| 27 | Pollastri et al., Improving skin lesion segmentation with generative adversarial networks | 2018 | peer-reviewed conference | ISIC2017 | - | Jaccard, l1 | 78.10% | ✗ | GAN | ✓ | ✗ |
| 50 | Cui et al., Ensemble Transductive Learning for Skin Lesion Segmentation | 2019 | peer-reviewed conference | ISIC2018 | dilated conv., parallel m.s. conv., separable conv. | - | 83.00% | ✗ | - | ✗ | ✗ |
| 79 | Al Nazi and Abir, Automatic Skin Lesion Segmentation and Melanoma Detection: Transfer Learning Approach with U-Net and DCNN-SVM | 2020 | peer-reviewed conference | ISIC2018, PH2 | skip con. | Dice | 80.00% | ✓ | rotation, zooming, flipping, elastic dist., Gaussian dist., histogram equal., color jittering | ✗ | ✓ |
| 51 | Song et al., Dense-Residual Attention Network for Skin Lesion Segmentation | 2019 | peer-reviewed conference | ISIC2017 | skip con., residual con., dense con., attention mod. | CE, Jaccard | 76.50% | ✗ | - | ✗ | ✗ |
| 52 | Singh et al., FCA-Net: Adversarial Learning for Skin Lesion Segmentation Based on Multi-Scale Features and Factorized Channel Attention | 2019 | peer-reviewed journal | ISIC2016, ISIC2017, ISIC2018 | skip con., residual con., factorized conv., attention mod., GAN | CE, l1, EPE | 78.65% | ✗ | - | ✗ | ✓ |
| 53 | Tan et al., Evolving ensemble models for image segmentation using enhanced particle swarm optimization | 2019 | peer-reviewed journal | ISIC2017, DermoFit, PH2 | dilated conv. | Dice | 62.29%* | ✓ | - | ✓ | ✗ |
| 54 | Kaul et al., FocusNet: an attention-based fully convolutional network for medical image segmentation | 2019 | peer-reviewed conference | ISIC2017 | skip con., residual con., attention mod. | Dice | 75.60% | ✗ | channel shift | ✗ | ✗ |
| 55 | De Angelo et al., Skin lesion segmentation using deep learning for images acquired from smartphones | 2019 | peer-reviewed conference | ISIC2017, Private | skip con. | CE, Dice | 76.07% | ✗ | flipping, shifting, rotation, color jittering | ✓ | ✗ |
| 56 | Zhang et al., DSM: A Deep Supervised Multi-Scale Network Learning for Skin Cancer Segmentation | 2019 | peer-reviewed journal | ISIC2017, PH2 | skip con., residual con., parallel m.s. conv. | CE, Dice, DS | 78.50% | ✓ | flipping, rotation, whitening, contrast enhance. | ✓ | ✗ |
| 28 | Vesal et al., SkinNet: A deep learning framework for skin lesion segmentation | 2018 | abstract | ISIC2017 | dilated conv., dense con., skip con. | Dice | 76.67% | ✗ | rotation, flipping, translation, scaling, color shift | ✗ | ✗ |
| 57 | Soudani and Barhoumi, An image-based segmentation recommender using crowdsourcing and transfer learning for skin lesion extraction | 2019 | peer-reviewed journal | ISIC2017 | residual con. | CE | 78.60% | ✗ | rotation, flipping | ✗ | ✗ |
| 58 | Mirikharaji et al., Learning to segment skin lesions from noisy annotations | 2019 | peer-reviewed conference | ISIC2017 | skip con. | WCE | 68.91%* | ✗ | - | ✗ | ✗ |
| 29 | Chen et al., A multi-task framework with feature passing module for skin lesion classification and segmentation | 2018 | peer-reviewed conference | ISIC2017 | residual con., dilated conv., parallel m.s. conv. | WCE | 78.70% | ✗ | rotation, flipping, cropping, zooming, Gaussian noise | ✓ | ✗ |
| 59 | Nasr-Esfahani et al., Dense pooling layers in fully convolutional network for skin lesion segmentation | 2019 | peer-reviewed journal | DermQuest | dense con., | WCE | 85.20% | ✗ | rotation, flipping, cropping | ✗ | ✗ |
| 30 | Jahanifar et al., Segmentation of skin lesions and their attributes using multi-scale convolutional neural networks and domain specific augmentations | 2018 | non peer-reviewed technical report | ISIC2016, ISIC2017, ISIC2018 | skip con., pyramid pooling, parallel m.s. conv. | Tanimoto | 80.60% | ✓ | flipping, rotation, zooming, translation, shearing, color shift, intensity scaling, adding noises, contrast adjust., sharpness adjust., disturb illumination, hair occlusion | ✓ | ✗ |
| 60 | Wang et al., Automated Segmentation of Skin Lesion Based on Pyramid Attention Network | 2019 | peer-reviewed conference | ISIC2017, ISIC2018 | skip con., residual con., parallel m.s. conv., attention mod. | WDice | 77.60% | ✗ | copping, flipping | ✗ | ✗ |
| 61 | Sarker et al., MobileGAN: Skin Lesion Segmentation Using a Lightweight Generative Adversarial Network | 2019 | non peer-reviewed technical report | ISIC2017, ISIC2018 | factrized conv., attention mod., GAN | CE, Jaccard, l1, ADV | 77.98% | ✗ | flipping, gamma reconst., contrast adjust. | ✗ | ✗ |
| 31 | Mirikharaji et al., Deep auto-context fully convolutional neural network for skin lesion segmentation | 2018 | peer-reviewed conference | ISIC2016 | skip con. | CE | 83.30% | ✗ | flipping, rottaion | ✗ | ✗ |
| 62 | Tu et al., Dense-Residual Network With Adversarial Learning for Skin Lesion Segmentation | 2019 | peer-reviewed journal | ISIC2017, PH2 | skip con., residual con., dense con., GAN | Jaccard, EPE, l1, DS, ADV | 76.80% | ✓ | flipping | ✗ | ✗ |
| 63 | Wei et al., Attention-Based DenseUnet Network With Adversarial Training for Skin Lesion Segmentation | 2019 | peer-reviewed journal | ISIC2016, ISIC2017, PH2 | skip con., residual con., attention mod., GAN | Jaccard, l1, ADV | 80.45% | ✓ | rotation, flipping, color jittering | ✗ | ✗ |
| 64 | Ünver and Ayan, Skin lesion segmentation in dermoscopic images with combination of YOLO and GrabCut algorithm | 2019 | peer-reviewed journal | ISIC2017, PH2 | - | l2 | 74.81% | ✓ | - | ✓ | ✗ |
| 65 | Al-masni et al., A Deep Learning Model Integrating FrCN and Residual Convolutional Networks for Skin Lesion Segmentation and Classification | 2019 | peer-reviewed conference | ISIC2017 | - | - | 77.11% | ✗ | rotation, flipping | ✗ | ✗ |
| 80 | Deng et al., Weakly and Semi-supervised Deep Level Set Network for Automated Skin Lesion Segmentation | 2020 | peer-reviewed conference | ISIC2017, PH2 | dilated conv., parallel m.s. conv., separable conv., semi-supervised | Dice, Narrowband suppression | 83.9% | ✓ | rotation | ✓ | ✗ |
| 66 | Canalini et al., Skin lesion segmentation ensemble with diverse training strategies | 2019 | peer-reviewed conference | ISIC2017 | dilated conv., parallel m.s. conv., separable conv. | CE, Tanimoto | 85.00% | ✗ | rotating, flipping, shifting, shearing, scaling, color jittering | ✓ | ✗ |
| 67 | Wang et al., Dermoscopic Image Segmentation Through the Enhanced High-Level Parsing and Class Weighted Loss | 2019 | peer-reviewed conference | ISIC2017 | residual con. | WCE | 78.10% | ✗ | flipping, scaling | ✗ | ✗ |
| 68 | Alom et al., Skin cancer segmentation and classification with improved deep convolutional neural network | 2020 | peer-reviewed conference | ISIC2018 | skip con., residual con., recurrent CNN | CE | 88.83% | ✗ | flipping | ✗ | ✗ |
| 69 | Pollastri et al., Augmenting Data with GANs to Segment Melanoma Skin Lesions | 2020 | peer-reviewed journal | ISIC2017 | - | Tanimoto | 78.90% | ✗ | GAN, flipping, rotation, shifting, scaling, color jittering | ✗ | ✗ |
| 70 | Liu et al., Skin Lesion Segmentation Based on Improved U-Net | 2019 | peer-reviewed conference | ISIC2017 | skip con., dilated conv. | CE | 75.20% | ✗ | scaling, cropping, rotation, flipping, image deformation | ✗ | ✗ |
| 4 | Ramachandram and Taylor, Skin lesion segmentation using deep hypercolumn descriptors | 2017 | peer-reviewed journal | ISIC2017 | - | CE | 79.20% | ✗ | rotation, flipping, color jittering | ✗ | ✗ |
| 32 | Bi et al., Improving automatic skin lesion segmentation using adversarial learning based data augmentation | 2018 | non peer-reviewed technical report | ISIC2018 | residual con. | CE | 83.12% | ✗ | GAN | ✗ | ✗ |
| 71 | Abhishek and Hamarneh, Mask2Lesion: Mask-constrained adversarial skin lesion image synthesis | 2019 | peer-reviewed conference | ISIC2017, PH2 | skip con. | - | 68.69%* | ✓ | rotation, flipping, GAN | ✗ | ✓ |
| 81 | Xie et al., A Mutual Bootstrapping Model for Automated Skin Lesion Segmentation and Classification | 2020 | peer-reviewed journal | ISIC2017, PH2 | dilated conv., parallel m.s. conv., separable conv. | Dice, Rank | 80.4% | ✓ | cropping, scaling, rotation, shearing, shifting, zooming, whitening, flipping | ✗ | ✓ |
| 33 | He et al., Dense deconvolution net: Multi path fusion and dense deconvolution for high resolution skin lesion segmentation | 2018 | peer-reviewed journal | ISIC2016, ISIC2017 | skip con., residual con., image pyramid | CE, Dice, DS | 76.10% | ✗ | rotation | ✓ | ✗ |
| 5 | Yu et al., Automated melanoma recognition in dermoscopy images via very deep residual networks | 2017 | peer-reviewed journal | ISIC2016 | skip con., residual con. | - | 82.90% | ✗ | rotation, translation, random noise, cropping | ✗ | ✓ |
| 6 | Bi et al., Dermoscopic image segmentation via multistage fully convolutional networks | 2017 | peer-reviewed journal | ISIC2016, PH2 | - | CE | 84.64% | ✓ | flipping, cropping | ✓ | ✗ |
| 1 | Jafari et al., Skin lesion segmentation in clinical images using deep learning | 2016 | peer-reviewed conference | DermQuest | image pyramid | - | - | ✗ | - | ✓ | ✗ |
| 7 | Jafari et al., Extraction of skin lesions from non-dermoscopic images for surgical excision of melanoma | 2017 | peer-reviewed journal | DermQuest | image pyramid | - | - | ✗ | - | ✓ | ✗ |
| 34 | Xue et al., Adversarial learning with multi-scale loss for skin lesion segmentation | 2018 | peer-reviewed conference | ISIC2017 | skip con., residual con., global conv., GAN | l1, DS, ADV | 78.50% | ✗ | cropping, color jittering | ✗ | ✗ |
| 8 | Yuan et al., Automatic skin lesion segmentation using deep fully convolutional networks with jaccard distance | 2017 | peer-reviewed journal | ISIC2016, PH2 | - | Tanimoto | 84.7% | ✓ | flipping, rotation, scaling, shifting, contrast norm. | ✓ | ✗ |
| 82 | Zhang et al., Kappa loss for skin lesion segmentation in fully convolutional network | 2020 | peer-reviewed conference | SCD, ISIC2016, ISIC2017, ISIC2018 | skip con. | Kappa Loss | 84.00%* | ✗ | rotation, shifting, shearing, zooming, flipping | ✗ | ✓ |
| 83 | Saha et al., Leveraging adaptive color augmentation in convolutional neural networks for deep skin lesion segmentation | 2020 | peer-reviewed conference | ISIC2017, ISIC2018 | skip con., dense con. | CE | 81.9% | ✗ | color jittering, rotation, flipping, translation | ✗ | ✗ |
| 84 | Henry et al., MixModule: Mixed CNN Kernel Module for Medical Image Segmentation | 2020 | peer-reviewed conference | ISIC2018 | skip con., parallel m. s. conv., attention mod. | - | 78.04% | ✗ | color jittering, rotation, cropping, flipping, shift | ✗ | ✓ |
| 85 | Jafari et al., DRU-Net: An Efficient Deep Convolutional Neural Network for Medical Image Segmentation | 2020 | peer-reviewed conference | ISIC2018 | skip con., residual con., dense con. | CE | 75.5% | ✗ | - | ✗ | ✓ |
| 86 | Li et al., A generic ensemble based deep convolutional neural network for semi-supervised medical image segmentation | 2020 | peer-reviewed conference | ISIC2018 | skip con., residual con., ensemble, semi-supervised | CE, Dice | 75.5% | ✗ | - | ✗ | ✗ |
| 87 | Guo et al., Complementary network with adaptive receptive fields for melanoma segmentation | 2020 | peer-reviewed conference | ISIC2018 | skip con., dilated conv., parallel m. s. conv. | Focal, Jaccard | 77.60% | ✗ | - | ✗ | ✓ |
| 88 | Li et al., A multi-task self-supervised learning framework for scopy images | 2020 | peer-reviewed conference | ISIC2018 | skip con., residual con., self-supervised | MSE, KL div. | 87.74%* | ✗ | - | ✗ | ✗ |
| 9 | Ramachandram and DeVries, LesionSeg: semantic segmentation of skin lesions using deep convolutional neural network | 2017 | non peer-reviewed technical report | ISIC2017 | dilated conv. | CE | 64.20% | ✗ | rotation, flipping | ✓ | ✗ |
| 10 | Bozorgtabar et al., Investigating deep side layers for skin lesion segmentation | 2017 | peer-reviewed conference | ISIC2016 | - | CE | 82.90% | ✗ | rotations | ✓ | ✗ |
| 11 | Bi et al., Semi-automatic skin lesion segmentation via fully convolutional networks | 2017 | peer-reviewed conference | ISIC2016 | parallel m. s. | - | 86.36% | ✗ | crops, flipping | ✓ | ✗ |
| 12 | Attia et al., Skin melanoma segmentation using recurrent and convolutional neural networks | 2017 | peer-reviewed conference | ISIC2016 | recurrent net. | - | 93.00% | ✗ | - | ✗ | ✗ |
| 13 | Deng et al., Segmentation of dermoscopy images based on fully convolutional neural network | 2017 | peer-reviewed conference | ISIC2016 | parallel m. s. | - | 84.1% | ✗ | - | ✗ | ✗ |
| 14 | Mishra and Daescu, Deep learning for skin lesion segmentation | 2017 | peer-reviewed conference | ISIC2017 | skip con. | Dice | 84.2% | ✗ | rotation, flipping | ✓ | ✗ |
| 15 | Goyal et al., Multi-class semantic segmentation of skin lesions via fully convolutional networks | 2017 | peer-reviewed conference | ISIC2017 | - | CE, Dice | - | ✗ | - | ✗ | ✗ |
| 89 | Jiang et al., Skin Lesion Segmentation Based on Multi-Scale Attention Convolutional Neural Network | 2020 | peer-reviewed journal | ISIC2017, PH2 | skip con., residual con., attention mod. | CE | 73.35% | ✗ | flipping | ✗ | ✗ |
| 90 | Qiu et al., Inferring Skin Lesion Deep Convolutional Neural Networks | 2020 | peer-reviewed journal | ISIC2017, PH2 | ensemble | - | 80.02% | ✗ | translation, rotation, shearing | ✓ | ✗ |
| 91 | Xie et al., Skin lesion segmentation using high-resolution convolutional neural network | 2020 | peer-reviewed journal | ISIC2016, ISIC2017, PH2 | attention mod. | CE | 78.3% | ✗ | rotation, flipping | ✗ | ✗ |
| 92 | Zafar et al., Skin lesion segmentation from dermoscopic images using convolutional neural network | 2020 | peer-reviewed journal | ISIC2017, PH2 | skip con., residual con. | CE | 77.2% | ✗ | rotation | ✗ | ✗ |
| 72 | Shahin et al., Deep convolutional encoder-decoders with aggregated multi-resolution skip connections for skin lesion segmentation | 2019 | peer-reviewed conference | ISIC2018 | skip con., image pyramid | Generalized, Dice | 73.8% | ✗ | rotation, flipping, zooming | ✗ | ✗ |
| 73 | Adegun and Viriri, An enhanced deep learning framework for skin lesions segmentation | 2019 | peer-reviewed conference | ISIC2017 | - | Dice | 83.0% | ✗ | elastic | ✗ | ✗ |
| 93 | Azad et al., Attention Deeplabv3+: Multi-level Context Attention Mechanism for Skin Lesion Segmentation | 2020 | peer-reviewed conference | ISIC 2017, ISIC 2018, PH2 | dilated conv., attention mod. | - | 96.98% | ✗ | - | ✗ | ✓ |
| 94 | Nathan and Kansal, Lesion Net--Skin Lesion Segmentation Using Coordinate Convolution and Deep Residual Units | 2020 | non peer-reviewed technical report | ISIC 2016, ISIC 2017, ISIC 2018, PH2 | skip con., residual con. | CE, Dice | 78.28% | ✗ | rotation, flipping, shearing, zoom | ✗ | ✗ |
| 95 | Mirikharaji et al., D-LEMA: Deep Learning Ensembles from Multiple Annotations-Application to Skin Lesion Segmentation | 2021 | peer-reviewed conference | ISIC Archive, PH2, DermoFit | skip con., residual con., ensemble | CE | 72.11% | ✗ | - | ✗ | ✗ |
| 109 | Arora et al., Automated skin lesion segmentation using attention-based deep convolutional neural network | 2021 | peer-reviewed journal | ISIC 2018 | skip con., attention mod. | Dice, Tversky, Focal Tversky | 83% | ✗ | flipping | ✓ | ✗ |
| 74 | Taghanaki et al., Improved inference via deep input transfer | 2019 | peer-reviewed conference | ISIC 2017 | skip con. | Dice, l1, SSIM | 69.35%* | ✗ | rotation, flipping, gradient-based, perturbation | ✗ | ✗ |
| 96 | Öztürk and Özkaya, Skin lesion segmentation with improved convolutional neural network | 2020 | peer-reviewed journal | ISIC 2017, PH2 | residual con. | - | 78.34% | ✓ | - | ✗ | ✗ |
| 97 | Abhishek et al., Illumination-based Transformations Improve Skin Lesion Segmentation in Dermoscopic Images | 2020 | peer-reviewed conference | ISIC 2017, DermoFit, PH2 | skip con. | Dice | 75.70% | ✓ | rotation, flipping | ✗ | ✓ |
| 75 | Saini et al., Detector-SegMentor Network for Skin Lesion Localization and Segmentation | 2019 | peer-reviewed conference | ISIC 2017, ISIC 2018, PH2 | skip con., multi-task | Dice | 84.9% | ✗ | rotation, flipping, shearing, stretch, crop, contrast | ✗ | ✗ |
| 98 | Kaymak et al., Skin lesion segmentation using fully convolutional networks: A comparative experimental study | 2020 | peer-reviewed journal | ISIC 2017 | - | - | 72.5% | ✗ | - | ✗ | ✗ |
| 99 | Bagheri et al., Two-stage Skin Lesion Segmentation from Dermoscopic Images by Using Deep Neural Networks | 2020 | peer-reviewed journal | ISIC2017, DermQuest | dilated conv., parallel m.s. conv., separable conv. | - | 79.05% | ✓ | rotation, flipping, brightness change, resizing | ✗ | ✗ |
| 100 | Jayapriya and Jacob, Hybrid fully convolutional networks-based skin lesion segmentation and melanoma detection using deep feature | 2020 | peer-reviewed journal | ISIC2016 | skip con., parallel m.s. conv. | - | 92.42% | ✗ | - | ✗ | ✗ |
| 101 | Wang et al., Cascaded Context Enhancement for Automated Skin Lesion Segmentation | 2020 | non peer-reviewed technical report | ISIC2016, ISIC2017, PH2 | residual con., dilated conv., attention mod. | CE, Dice, DS | 80.30% | ✓ | flipping, rotation, cropping | ✗ | ✗ |
| 76 | Wang et al., Bi-directional dermoscopic feature learning and multi-scale consistent decision fusion for skin lesion segmentation | 2019 | peer-reviewed journal | ISIC2016, ISIC2017 | skip con., residual con., dilated conv. | WCE | 81.47% | ✗ | flipping, scaling | ✗ | ✗ |
| 110 | Jin et al., Cascade knowledge diffusion network for skin lesion diagnosis and segmentation | 2021 | peer-reviewed journal | ISIC2017, ISIC2018 | skip con., residual con., attention mod. | Dice, Focal | 80.00% | ✗ | flipping, rotation, affine trans., scaling, cropping | ✗ | ✓ |
| 111 | Hasan et al., Dermo-DOCTOR: A framework for concurrent skin lesion detection and recognition using a deep convolutional neural network with end-to-end dual encoders | 2021 | peer-reviewed journal | ISIC 2016, ISIC 2017 | skip con., residual con., separable conv. | Dice, CE | 66.66%* | ✗ | flipping, rotation, shifting, zooming, intensity adjust. | ✗ | ✗ |
| 112 | Kosgiker et al., SegCaps: An efficient SegCaps network-based skin lesion segmentation in dermoscopic images | 2021 | peer-reviewed journal | ISIC 2017, PH2 | - | MSE, CE | 90.25% | ✗ | - | ✗ | ✗ |
| 102 | Wang et al., DONet: Dual Objective Networks for Skin Lesion Segmentation | 2020 | non peer-reviewed technical report | ISIC2018, PH2 | attention mod., skip con., parallel m.s. conv., recurrent CNN, | Dice, Focal Tversky | 80.6% | ✗ | rotation, flipping, cropping | ✗ | ✗ |
| 103 | Ribeiro et al., Less is more: Sample selection and label conditioning improve skin lesion segmentation | 2020 | peer-reviewed conference | ISIC Archive, PH2, DermoFit | skip con., residual con., dilated conv. | Soft Jaccard, CE | - | ✓ | Gaussian noise, color jittering | ✓ | ✓ |
| 35 | Ebenezer and Rajapakse, Automatic segmentation of skin lesions using deep learning | 2018 | non peer-reviewed technical report | ISIC 2018 | skip con. | Dice | 75.6% | ✗ | rotation, flipping, zooming | ✓ | ✓ |
| 113 | Bagheri et al., Skin lesion segmentation based on mask RCNN, Multi Atrous Full-CNN, and a geodesic method | 2021 | peer-reviewed journal | ISIC2016, ISIC2017, ISIC2018, PH2, DermQuest | parallel m.s. conv., dilated conv. | Dice, CE | 85.04% | ✓ | rotation, flipping, color jittering | ✗ | ✗ |
| 114 | Saini et al., B-SegNet: branched-SegMentor network for skin lesion segmentation | 2021 | peer-reviewed conference | ISIC2017, ISIC2018, PH2 | pyramid pooling, residual con., skip con., dilated conv., attention mod. | Dice | 85.00% | ✓ | rotation, shearing, color jittering | ✗ | ✗ |
| 115 | Tong et al., ASCU-Net: Attention Gate, Spatial and Channel Attention U-Net for Skin Lesion Segmentation | 2021 | peer-reviewed journal | ISIC2016, ISIC2017, PH2 | skip con., attention mod. | CE | 84.2% | ✓ | flipping | ✗ | ✗ |
| 116 | Bagheri et al., Skin lesion segmentation from dermoscopic images by using Mask R-CNN, Retina-Deeplab, and graph-based methods | 2021 | peer-reviewed journal | DermQuest, ISIC2017, PH2 | ensemble | CE, Focal | 86.53% | ✓ | rotation, flipping, color jittering | ✓ | ✗ |
| 117 | Ren et al., Serial attention network for skin lesion segmentation | 2021 | peer-reviewed journal | ISIC2017 | dense con., dilated conv., separable conv., attention mod. | Dice, CE | 76.92% | ✗ | flipping, rotation | ✗ | ✗ |
| 118 | Liu et al., Skin lesion segmentation using deep learning with auxiliary task | 2021 | peer-reviewed journal | ISIC2017 | residual con., dilated conv., pyramid pooling | WCE | 79.46% | ✗ | flipping, cropping, rotation, image deformation | ✗ | ✗ |
| 119 | Khan et al., PMED-Net: Pyramid Based Multi-Scale Encoder-Decoder Network for Medical Image Segmentation | 2021 | peer-reviewed journal | ISIC2018 | skip con., image pyramid | Dice | 85.10% | ✗ | - | ✗ | ✓ |
| 77 | Kamalakannan et al., Self-Learning AI Framework for Skin Lesion Image Segmentation and Classification | 2019 | peer-reviewed journal | ISIC Archive | skip con. | CE | - | ✗ | - | ✗ | ✗ |
| 104 | Zhu et al., ASNet: An adaptive scale network for skin lesion segmentation in dermoscopy images | 2020 | peer-reviewed conference | ISIC2018 | skip con., residual con., dilated conv., attention mod. | CE, Dice | 82.15% | ✗ | flipping | ✗ | ✗ |
| 120 | Redekop and Chernyavskiy, Uncertainty-based method for improving poorly labeled segmentation datasets | 2021 | peer-reviewed conference | ISIC2017 | - | - | 68.77%* | ✗ | - | ✗ | ✗ |
| 121 | Kaul et al., Focusnet++: Attentive Aggregated Transformations For Efficient And Accurate Medical Image Segmentation | 2021 | peer-reviewed conference | ISIC2018 | skip con., residual con., attention mod. | CE, Tversky, adaptive, logarithmic | 82.71% | ✗ | - | ✗ | ✓ |
| 122 | Abhishek and Hamarneh, Matthews correlation coefficient loss for deep convolutional networks: Application to skin lesion segmentation | 2021 | peer-reviewed conference | ISIC2017, PH2, DermoFit | skip con. | MCC | 75.18% | ✗ | flipping, rotation | ✗ | ✓ |
| 123 | Tang et al., Introducing frequency representation into convolution neural networks for medical image segmentation via twin-Kernel Fourier convolution | 2021 | peer-reviewed journal | ISIC2018 | skip con. | CE | 78.25% | ✗ | - | ✗ | ✗ |
| 124 | Xie et al., Semi-Supervised Skin Lesion Segmentation with Learning Model Confidence | 2021 | peer-reviewed conference | ISIC2018 | dilated conv., semi-supervised | CE, KL div. | 82.37% | ✗ | scaling, rotation, elastic transformation | ✗ | ✗ |
| 125 | Poudel and Lee, Deep multi-scale attentional features for medical image segmentation | 2021 | peer-reviewed journal | ISIC2017 | skip con., attention mod. | CE | 87.44% | ✗ | scaling, flipping, rotation, Gaussian noise, median blur | ✗ | ✗ |
| 126 | Şahin et al., Robust optimization of SegNet hyperparameters for skin lesion segmentation | 2021 | peer-reviewed journal | ISIC2016, ISIC 2017 | skip con., Gaussian process | - | 74.51% | ✗ | resize, rotation, reflection | ✓ | ✗ |
| 127 | Sarker et al., SLSNet: Skin lesion segmentation using a lightweight generative adversarial network | 2021 | peer-reviewed journal | ISIC 2017, ISIC 2018 | parallel m.s. conv., attention mod., GAN | l1, Jaccard | 81.98% | ✗ | flipping, contrast, gamma reconstruction | ✗ | ✗ |
| 128 | Wang et al., Knowledge-aware Deep Framework for Collaborative Skin Lesion Segmentation and Melanoma Recognition | 2021 | peer-reviewed journal | ISIC 2016, ISIC 2017 | residual con., skip con., lesion-based pooling, feature fusion | CE | 82.4% | ✗ | flipping, scaling, cropping | ✗ | ✗ |
| 129 | Sachin et al., Performance Analysis of Deep Learning Models for Biomedical Image Segmentation | 2021 | book chapter | ISIC 2018 | residual con., skip con. | - | 75.96% | ✗ | flipping, scaling, color jittering | ✗ | ✗ |
| 130 | Wibowo et al., Lightweight encoder-decoder model for automatic skin lesion segmentation | 2021 | peer-reviewed journal | ISIC 2017, ISIC 2018, PH2 | BConvLSTM, separable conv., residual con., skip con. | Jaccard | 80.25% | ✗ | distortion, blur, color jittering, contrast, gamma sharpen | ✓ | ✓ |
| 131 | Gudhe et al., Multi-level dilated residual network for biomedical image segmentation | 2021 | peer-reviewed journal | ISIC 2018 | dilated conv., residual con., skip con. | CE | 91% | ✗ | flipping, scaling, shearing, color jittering, Gaussian blur, Gaussian noise | ✗ | ✓ |
| 132 | Khouloud et al., W-net and inception residual network for skin lesion segmentation and classification | 2021 | peer-reviewed journal | ISIC 2016, ISIC 2017, ISIC 2018, PH2 | feature pyramid, residual con., skip con., attention mod. | - | 86.92%* | ✗ | - | ✗ | ✗ |
| 105 | Gu et al., CA-Net: Comprehensive attention convolutional neural networks for explainable medical image segmentation | 2020 | peer-reviewed journal | ISIC 2018 | residual con., skip con., attention mod. | Dice | 85.32%* | ✗ | cropping, flipping, rotation | ✗ | ✓ |
| 133 | Gu et al., kCBAC-Net: Deeply Supervised Complete Bipartite Networks with Asymmetric Convolutions for Medical Image Segmentation | 2021 | peer-reviewed conference | ISIC 2017 | asymmetric conv., skip con. | DS | 79.4% | ✗ | cropping, flipping, rotation | ✗ | ✗ |
| 134 | Zhao et al., Segmentation of dermoscopy images based on deformable 3D convolution and ResU-NeXt++ | 2021 | peer-reviewed journal | ISIC 2018 | pyramid pooling, attention mod., residual con., skip con. | CE, Dice | 86.84% | ✗ | cropping | ✗ | ✗ |
| 135 | Tang et al., AFLN-DGCL: Adaptive Feature Learning Network with Difficulty-Guided Curriculum Learning for skin lesion segmentation | 2021 | peer-reviewed journal | ISIC 2016, ISIC 2017, ISIC 2018 | attention mod., residual con., skip con., ensemble, pyramid pooling | Focal | 80.7% | ✗ | copying | ✗ | ✗ |
| 136 | Zunair and Hamza, Sharp U-Net: Depthwise convolutional network for biomedical image segmentation | 2021 | peer-reviewed journal | ISIC 2018 | sharpening kernel, residual con. | CE | 79.78% | ✗ | - | ✗ | ✓ |
| 147 | Dai et al., Ms RED: A novel multi-scale residual encoding and decoding network for skin lesion segmentation | 2022 | peer-reviewed journal | ISIC2018, PH2 | residual con., skip con., dilated conv., image pyramid, attention mod. | CE, Dice, SoftDice | 83.45% | ✓ | cropping, flipping, rotation | ✗ | ✗ |
| 148 | Bi et al., Hyper-fusion network for semi-automatic segmentation of skin lesions | 2022 | peer-reviewed journal | ISIC2016, ISIC2017, PH2 | residual con., skip con., attention mod., feature fusion | CE | 83.70% | ✓ | cropping, flipping | ✗ | ✗ |
| 137 | Li et al., Superpixel-Guided Iterative Learning from Noisy Labels for Medical Image Segmentation | 2021 | peer-reviewed conference | ISIC 2017 | skip con. | CE, KL div. | 71.12%* | ✗ | - | ✗ | ✓ |
| 138 | Zhang et al., Self-supervised Correction Learning for Semi-supervised Biomedical Image Segmentation | 2021 | peer-reviewed conference | ISIC 2016 | skip con., residual con., feature fusion, semi-supervised, self-supervised | CE, Dice | 80.49% | ✗ | flipping, rotation, zooming, cropping | ✗ | ✓ |
| 139 | Xu et al., DC-Net: Dual context network for 2D medical image segmentation | 2021 | peer-reviewed conference | ISIC 2018 | Transformer, multi-scale | Dice | 89.6% | ✗ | flipping, rotation | ✗ | ✗ |
| 140 | Ahn et al., A spatial guided self-supervised clustering network for medical image segmentation | 2021 | peer-reviewed conference | PH2 | self-supervised, clustering | CE, Spatial loss, Consistency loss | 71.53%* | ✗ | - | ✗ | ✓ |
| 141 | Zhang et al., TransFuse: Fusing Transformers and CNNs for medical image segmentation | 2021 | peer-reviewed conference | ISIC 2017 | skip con., feature fusion, Transformer | CE, Jaccard | 79.5% | ✗ | rotation, flipping, color jittering | ✗ | ✓ |
| 142 | Ji et al., Multi-compound Transformer for accurate biomedical image segmentation | 2021 | peer-reviewed conference | ISIC 2018 | skip con., multi-scale, Transformer | CE, Dice | 82.4%* | ✗ | flipping | ✗ | ✓ |
| 143 | Wang et al., Boundary-aware Transformers for skin lesion segmentation | 2021 | peer-reviewed conference | ISIC 2016, ISIC 2018, PH2 | multi-scale, Transformer | CE, Dice | 84.3%* | ✓ | flipping, scaling | ✗ | ✓ |
| 149 | Lin et al., ConTrans: Improving Transformer with Convolutional Attention for Medical Image Segmentation | 2022 | peer-reviewed conference | ISIC 2017, ISIC 2018 | attention mod., Transformer | CE, Jaccard, DS | 77.81%* | ✗ | flipping, rotation | ✗ | ✗ |
| 150 | Wu et al., SeATrans: Learning Segmentation-Assisted Diagnosis Model via Transformer | 2022 | peer-reviewed conference | PH2 | skip con., Transformer, multi-scale | CE | 70.0%* | ✗ | - | ✗ | ✗ |
| 151 | Valanarasu and Patel, UNeXt: MLP-based Rapid Medical Image Segmentation Network | 2022 | peer-reviewed conference | ISIC 2018 | skip con. | CE, Dice | 81.7% | ✗ | - | ✗ | ✓ |
| 152 | Basak et al., MFSNet: A multi focus segmentation network for skin lesion segmentation | 2022 | peer-reviewed journal | ISIC 2017, PH2, HAM10000 | residual con., multi-scale, attention mod. | CE, Jaccard, DS | 97.4% | ✗ | - | ✗ | ✓ |
| 153 | Wu et al., FAT-Net: Feature adaptive Transformers for automated skin lesion segmentation | 2022 | peer-reviewed journal | ISIC 2016, ISIC 2017, ISIC 2018, PH2 | skip con., residual con., attention mod., Transformer | CE, Dice | 76.53% | ✗ | flipping, rotation, brightness change, contrast change, change in H, S, V | ✗ | ✓ |
| 106 | Lei et al., Skin lesion segmentation via generative adversarial networks with dual discriminators | 2020 | peer-reviewed journal | ISIC 2017, ISIC 2018 | skip con., dense con., dilated conv., GAN | CE, l1, ADV | 77.1% | ✓ | flipping, rotation | ✗ | ✗ |
| 154 | Liu et al., NCRNet: Neighborhood Context Refinement Network for skin lesion segmentation | 2022 | peer-reviewed journal | ISIC 2017 | skip con., residual con., dilated conv., attention mod. | CE, Dice | 78.62% | ✗ | flipping, rotation | ✗ | ✗ |
| 155 | Wang et al., O-Net: a novel framework with deep fusion of CNN and Transformer for simultaneous segmentation and classification | 2022 | peer-reviewed journal | ISIC 2017 | skip con., residual con., Transformer | - | 84.52% | ✗ | flipping, rotation | ✗ | ✓ |
| 156 | Zhang et al., Feature Fusion for Segmentation and Classification of Skin Lesions | 2022 | peer-reviewed conference | ISIC 2017 | skip con., feature fusion | Dice, Focal | 74.54% | ✗ | flipping | ✗ | ✗ |
| 157 | Wang et al., Superpixel Inpainting For Self-Supervised Skin Lesion Segmentation from Dermoscopic Images | 2022 | peer-reviewed conference | ISIC 2017, PH2 | skip con., residual con., self-supervised | Dice | 76.5% | ✓ | rotation, flipping, color jittering | ✗ | ✗ |
| 144 | Yang et al., Deep Hybrid Convolutional Neural Network for Segmentation of Melanoma Skin Lesion | 2021 | peer-reviewed journal | ISIC 2018, PH2 | skip con., multi-scale, feature fusion | CE, Dice | 94.0% | ✗ | rotation, flipping, cropping, HSC, manipulation, luminance, and contrast shift | ✗ | ✗ |
| 107 | Andrade et al., Data Augmentation Using Adversarial Image-to-Image Translation for the Segmentation of Mobile-Acquired Dermatological Images | 2020 | peer-reviewed journal | DermoFit, SMARTSKINS | residual con., dilated conv., GAN | Dice | 81.03% | ✗ | flipping, brightness, saturation, contrast, hue, Gaussian hue | ✗ | ✗ |
| 145 | Tao et al., Attention-guided network with densely connected convolution for skin lesion segmentation | 2021 | peer-reviewed journal | ISIC 2017, PH2 | skip con., dense con., attention mod., multi-scale | - | 78.85% | ✗ | rotation | ✗ | ✗ |
| 108 | Wu et al., Automated skin lesion segmentation via an adaptive dual attention module | 2020 | peer-reviewed journal | ISIC 2017, ISIC 2018 | residual con., attention mod., multi-scale | CE, Dice | 82.55% | ✗ | flipping, rotation, scaling, cropping, sharpening, color, distribution adj., noise | ✗ | ✗ |
| 146 | Kim and Lee, A Simple Generic Method for Effective Boundary Extraction in Medical Image Segmentation | 2021 | peer-reviewed journal | ISIC 2016, PH2 | residual con., skip con. | boundary aware loss | 84.33%* | ✗ | - | ✗ | ✗ |
| 158 | Dong et al., TC-Net: Dual coding network of Transformer and CNN for skin lesion segmentation | 2022 | peer-reviewed journal | ISIC 2016, ISIC 2017, ISIC 2018 | residual con., skip con., Transformer, feature fusion | CE, Dice | 74.55% | ✗ | - | ✗ | ✗ |
| 159 | Chen et al., Skin Lesion Segmentation Using Recurrent Attentional Convolutional Networks | 2022 | peer-reviewed journal | ISIC 2017, PH2 | skip con., attention mod., recurrent net. | CE | 80.36% | ✓ | flipping, rotation, affine trans., masking, mesh distortion | ✗ | ✗ |
| 160 | Kaur et al., Automatic lesion segmentation using atrous convolutional deep neural networks in dermoscopic skin cancer images | 2022 | peer-reviewed journal | ISIC 2016, ISIC 2017, ISIC 2018, PH2 | dilated conv. | CE | 81.7% | ✓ | scaling, rotation, translation | ✗ | ✗ |
| 161 | Badshah and Ahmad, ResBCU-Net: Deep learning approach for segmentation of skin images | 2022 | peer-reviewed journal | ISIC 2018 | residual con., BConvLSTM | - | 94.5% | ✗ | - | ✗ | ✗ |
| 162 | Alam et al., S2C-DeLeNet: A parameter transfer based segmentation-classification integration for detecting skin cancer lesions from dermoscopic images | 2022 | peer-reviewed journal | HAM10000 | residual con., separable conv. | Dice | 91.1% | ✗ | - | ✗ | ✓ |
| 163 | Yu et al., mCA-Net: modified comprehensive attention convolutional neural network for skin lesion segmentation | 2022 | peer-reviewed journal | ISIC 2018 | skip con., attention mod., multi-scale | - | 87.89% | ✗ | - | ✗ | ✗ |
| 164 | Jiang et al., SEACU-Net: Attentive ConvLSTM U-Net with squeeze-and-excitation layer for skin lesion segmentation | 2022 | peer-reviewed journal | ISIC 2017, ISIC 2018 | skip con., attention mod., ConvLSTM | CE, Jaccard | 80.5% | ✗ | - | ✗ | ✗ |
| 165 | Ramadan et al., Color-invariant skin lesion semantic segmentation based on modified U-Net deep convolutional neural network | 2022 | peer-reviewed journal | ISIC 2018 | skip con., attention mod. | CE, Dice, sens.-spec. loss | 91.4% | ✗ | - | ✗ | ✗ |
| 166 | Zhang et al., Dynamic prototypical feature representation learning framework for semi-supervised skin lesion segmentation | 2022 | peer-reviewed journal | ISIC 2017, ISIC 2018 | skip con., dense con., semi-supervised | CE, contrastive loss | 73.89% | ✗ | scaling, flipping, color distortion | ✗ | ✗ |
| 167 | Tran and Pham, Fully convolutional neural network with attention gate and fuzzy active contour model for skin lesion segmentation | 2022 | peer-reviewed journal | ISIC 2017, PH2 | skip con., attention mod. | Focal Tversky, fuzzy loss | 79.2% | ✗ | rotation, zooming, flipping | ✗ | ✗ |
| 168 | Wang and Wang, Skin lesion segmentation with attention-based SC-Conv U-Net and feature map distortion | 2022 | peer-reviewed journal | ISIC 2017 | skip con., residual con., attention mod. | CE, Jaccard | 78.28% | ✗ | rotation, zooming, resizing, shifting | ✗ | ✗ |
| 169 | Zhao et al., Self-supervised Assisted Active Learning for Skin Lesion Segmentation | 2022 | peer-reviewed conference | ISIC 2017 | skip con., self-supervised | CE, Dice | 67.08%* | ✗ | - | ✗ | ✗ |
| 170 | Wang et al., Cross-Domain Few-Shot Learning for Rare-Disease Skin Lesion Segmentation | 2022 | peer-reviewed conference | PH2 | few shot, mask avg. pooling | Dice | 86.97%* | ✗ | - | ✗ | ✗ |
| 171 | Wang et al., CTCNet: A Bi-directional Cascaded Segmentation Network Combining Transformers with CNNs for Skin Lesions | 2022 | peer-reviewed conference | ISIC 2017, ISIC 2018 | residual con., dilated conv., multi-scale, feature fusion, Transformer | CE, Jaccard | 78.76% | ✗ | - | ✗ | ✗ |
| 172 | Liu et al., Skin Lesion Segmentation via Intensive Atrous Spatial Transformer | 2022 | peer-reviewed conference | ISIC 2017, ISIC 2018 | skip con., dilated conv., multi-scale, pyramid pooling, Transformer | CE | 80.19% | ✗ | - | ✗ | ✗ |
| 173 | Gu et al., DE-Net: A deep edge network with boundary information for automatic skin lesion segmentation | 2022 | peer-reviewed journal | ISIC 2017 | skip con., global adaptive, pooling | CE, l2 | 80.53% | ✗ | scaling, rotation, flipping | ✗ | ✗ |
| 174 | Khan et al., Ensemble learning of deep learning and traditional machine learning approaches for skin lesion segmentation and classification | 2022 | peer-reviewed journal | ISIC 2017, PH2 | residual con., attention mod., ensemble | CE | 79.2% | ✗ | - | ✗ | ✗ |
| 175 | Alahmadi and Alghamdi, Semi-Supervised Skin Lesion Segmentation With Coupling CNN and Transformer Features | 2022 | peer-reviewed journal | ISIC 2017, ISIC 2018, PH2 | skip con., feature fusion, semi-supervised, Transformer | CE, Dice, l2 | 82.78%* | ✗ | - | ✗ | ✗ |
| 176 | Li et al., MHAU-Net: Skin Lesion Segmentation Based on Multi-Scale Hybrid Residual Attention Network | 2022 | peer-reviewed journal | ISIC 2018 | skip con., residual con., dilated conv., attention mod., pyramid pooling, multi-scale | CE, Dice | 88.92% | ✗ | flipping, rotation | ✗ | ✗ |
| 177 | Kaur et al., Skin lesion segmentation using an improved framework of encoder-decoder based convolutional neural network | 2022 | peer-reviewed journal | ISIC 2016, ISIC 2017, ISIC 2018, PH2 | - | Tversky | 77.8% | ✓ | rotation, scaling | ✗ | ✗ |