Yanhang Shi
Skip Abstract Section
Abstract
In recent years, automatic skin lesion diagnosis methods based on artificial intelligence have achieved great success. However, the lack of labeled data, visual similarity between skin diseases, and restriction on private data sharing remain the major challenges in skin lesion diagnosis. In this article, first, we propose a federated contrastive learning framework to break down data silos and enhance the generalizability of diagnostic model to unseen data. Subsequently, by combining data features from different participated nodes, the proposed framework can improve the performance of contrastive training. To extract discriminative features during on-device training, we propose a contrastive learning based intelligent skin lesion diagnosis scheme in edge computing networks. Specifically, a contrastive learning based dual encoder network is designed to overcome training sample scarcity by fully leveraging unlabeled samples for performance improvement. Meanwhile, we devise a maximum mean discrepancy based supervised contrastive loss function, which can efficiently explore complex intra-class and inter-class variances of samples. Finally, the diagnosis simulations demonstrate that compared with existing methods, our proposed scheme can achieve superior accuracy in both on-device training and distributed training scenarios.
- [1] . 2021. Data augmentation for skin lesion using self-attention based progressive generative adversarial network. Expert Systems with Applications 165 (2021), 1–13.Google Scholar
Cross Ref
- [2] . 2019. Mask2Lesion: Mask-constrained adversarial skin lesion image synthesis. In Proceedings of Simulation and Synthesis in Medical Imaging: 4th International Workshop (SASHIMI’19). 71–80.Google ScholarDigital Library
- [3] . 2020. Integrated design of deep features fusion for localization and classification of skin cancer. Pattern Recognition Letters 131 (2020), 63–70.Google ScholarCross Ref
- [4] . 2018. Federated learning of predictive models from federated electronic health records. International Journal of Medical Informatics 112 (2018), 59–67.Google ScholarCross Ref
- [5] . 2019. Melanoma cancer classification using ResNet with data augmentation. In Proceedings of the 2nd International Seminar on Research of Information Technology and Intelligent Systems (ISRITI’19). 17–20.Google ScholarCross Ref
- [6] . 2019. Classification of skin lesions by combining multilevel learnings in a DenseNet architecture. In Proceedings of the 20th Image Analysis and Processing (ICIAP’19). 335–344.Google ScholarDigital Library
- [7] . 2020. A simple framework for contrastive learning of visual representations. In Proceedings of the 37th International Conference on Machine Learning (ICML’20). 1597–1607.Google ScholarDigital Library
- [8] . 2021. Federated contrastive learning for decentralized unlabeled medical images. In Proceedings of the 24th Medical Image Computing and Computer Assisted Intervention (MICCAI’21). 378–387.Google ScholarDigital Library
- [9] . 2021. A federated deep learning framework for 3D brain MRI images. In Proceedings of the International Joint Conference on Neural Networks (IJCNN’21). 1–6.Google ScholarCross Ref
- [10] . 2020. Bootstrap your own latent a new approach to self-supervised learning. In Proceedings of the 33rd International Conference on Neural Information Processing Systems (NIPS’20). 21271–21284.Google Scholar
- [11] . 2014. The global burden of skin disease in 2010: An analysis of the prevalence and impact of skin conditions. Journal of Investigative Dermatology 134, 6 (2014), 1527–1534.Google ScholarCross Ref
- [12] . 2020. Momentum contrast for unsupervised visual representation learning. In Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR’20). 9726–9735.Google ScholarCross Ref
- [13] . 2019. Mitochondrial organelle movement classification (fission and fusion) via convolutional neural network approach. IEEE Access 7 (2019), 86570–86577.Google ScholarCross Ref
- [14] . 2016. Skin lesion segmentation in clinical images using deep learning. In Proceedings of the 23rd International Conference on Pattern Recognition (ICPR’16). 337–342.Google ScholarCross Ref
- [15] . 2021. Federated semi-supervised learning with inter-client consistency & disjoint learning. In Proceedings of the 9th International Conference on Learning Representations (ICLR’21). 1–11.Google Scholar
- [16] . 2020. Skin lesions classification into eight classes for ISIC 2019 using deep convolutional neural network and transfer learning. IEEE Access 8 (2020), 114822–114832.Google ScholarCross Ref
- [17] . 2020. Supervised contrastive learning. In Proceedings of the 33rd Advances in Neural Information Processing Systems (NIPS’20), Vol. 33. 18661–18673.Google Scholar
- [18] . 2021. Blockchain-federated-learning and deep learning models for COVID-19 detection using CT imaging. IEEE Sensors Journal 21, 14 (2021), 16301–16314.Google ScholarCross Ref
- [19] . 2021. Model-contrastive federated learning. In Proceedings of the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR’21). 10708–10717.Google ScholarCross Ref
- [20] . 2020. Federated optimization in heterogeneous networks. In Proceedings of the 2nd Conference on Machine Learning and Systems (MLSys’20), Vol. 2. 429–450.Google Scholar
- [21] . 2015. Generative moment matching networks. In Proceedings of the 32nd International Conference on Machine Learning (ICML’15), Vol. 37. 1718–1727.Google Scholar
- [22] . 2020. Transfer learning using a multi-scale and multi-network ensemble for skin lesion classification. Computer Methods and Programs in Biomedicine 193 (2020), 105475.Google ScholarCross Ref
- [23] . 2019. Skin lesion classification using hybrid deep neural networks. In Proceedings of the 44th IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP’19). 1229–1233.Google ScholarCross Ref
- [24] . 2016. Combining deep learning and hand-crafted features for skin lesion classification. In Proceedings of the 6th International Conference on Image Processing Theory, Tools, and Applications (IPTA’16). 1–6.Google ScholarCross Ref
- [25] . 2017. Communication-efficient learning of deep networks from decentralized data. In Proceedings of the 20th International Conference on Artificial Intelligence and Statistics (AISTATS’17), Vol. 54. 1273–1282.Google Scholar
- [26] . 2019. Automated skin lesion classification using ensemble of deep neural networks in ISIC 2018: Skin lesion analysis towards melanoma detection challenge. arXiv:1901.10802 (2019).Google Scholar
- [27] . 2019. Risk-aware machine learning classifier for skin lesion diagnosis. Journal of Clinical Medicine 8 (2019), 1–24.Google ScholarCross Ref
- [28] . 2017. Transfer learning for melanoma detection: Participation in ISIC 2017 skin lesion classification challenge. arXiv:1703.05235 (2017).Google Scholar
- [29] . 2016. Deepmole: Deep neural networks for skin mole lesion classification. In Proceedings of the 23rd IEEE International Conference on Image Processing (ICIP’16). 2623–2627.Google ScholarCross Ref
- [30] . 2020. A GAN-based image synthesis method for skin lesion classification. Computer Methods and Programs in Biomedicine 195 (2020), 105568.Google ScholarCross Ref
- [31] . 2019. Skin lesion classification using GAN based data augmentation. In Proceedings of the 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC’19). 916–919.Google ScholarCross Ref
- [32] . 2018. Disease classification within dermascopic images using features extracted by ResNet50 and classification through deep forest. arXiv:1807.05711 (2018).Google Scholar
- [33] . 2017. Skin lesion classification from dermoscopic images using deep learning techniques. In Proceedings of the 13th IASTED International Conference on Biomedical Engineering (BioMed’17). 49–54.Google Scholar
- [34] . 2021. A patient-centric dataset of images and metadata for identifying melanomas using clinical context. Scientific Data 8 (2021), 1–8.Google Scholar
- [35] . 2019. Convolutional neural networks using MobileNet for skin lesion classification. In Proceedings of the 16th International Joint Conference on Computer Science and Software Engineering (JCSSE’19). 242–247.Google ScholarCross Ref
- [36] . 2021. Contrastive learning based intelligent skin lesion diagnosis in edge computing networks. In Proceedings of the 2021 IEEE Global Communications Conference (GLOBECOM’21). 1–6.Google ScholarDigital Library
- [37] . 2019. Federated learning in distributed medical databases: Meta-analysis of large-scale subcortical brain data. In Proceedings of the 16th International Symposium on Biomedical Imaging (ISBI’19). 270–274.Google ScholarCross Ref
- [38] . 2020. FixMatch: Simplifying semi-supervised learning with consistency and confidence. In Proceedings of the 33rd Advances in Neural Information Processing Systems (NIPS’20), Vol. 33. 596–608.Google Scholar
- [39] . 2017. Automatic skin lesion analysis towards melanoma detection. In Proceedings of the 21st Asia Pacific Symposium on Intelligent and Evolutionary Systems (IES’17). 106–111.Google ScholarCross Ref
- [40] . 2005. Assessing diagnostic skill in dermatology: A comparison between general practitioners and dermatologists. Australasian Journal of Dermatology 46 (2005), 230–234.Google ScholarCross Ref
- [41] . 2018. Representation learning with contrastive predictive coding. arXiv:1807.03748 (2018).Google Scholar
- [42] . 2021. Variation-aware federated learning with multi-source decentralized medical image data. IEEE Journal of Biomedical and Health Informatics 25, 7 (2021), 2615–2628.Google ScholarCross Ref
- [43] . 2021. MedMNIST classification decathlon: A lightweight AutoML benchmark for medical image analysis. In Proceedings of the 18th International Symposium on Biomedical Imaging (ISBI’21). 191–195.Google ScholarCross Ref
- [44] . 2020. Convolutional descriptors aggregation via cross-net for skin lesion recognition. Applied Soft Computing 92 (2020), 106281.Google ScholarCross Ref
- [45] . 2021. Dynamic-fusion-based federated learning for COVID-19 detection. IEEE Internet of Things Journal 8, 21 (2021), 15884–15891.Google ScholarCross Ref
- [46] . 2022. Divergence-aware federated self-supervised learning. In Proceedings of the 10th International Conference on Learning Representations (ICLR’22). 1–13.Google Scholar
Index Terms
- Skin Lesion Intelligent Diagnosis in Edge Computing Networks: An FCL Approach
Recommendations
A Computational Approach to Pertinent Feature Extraction for Diagnosis of Melanoma Skin Lesion
AbstractMelanoma, starts growing in melanocytes, is less common but more serious and aggressive than any other types of skin cancers found in human. Melanoma skin cancer can be completely curable if it is diagnosed and treated in an early stage. Biopsy is ...
Contrastive Learning Based Intelligent Skin Lesion Diagnosis in Edge Computing Networks
2021 IEEE Global Communications Conference (GLOBECOM)In recent years, automatic skin lesion diagnosis methods based on artificial intelligence (AI) have achieved great success. However, the lack of rich available training data and visual similarity of various skin diseases remain the major challenges in ...
Dynamic prototypical feature representation learning framework for semi-supervised skin lesion segmentation
AbstractAutomated skin lesion segmentation is an essential yet challenging task for computer-aided skin disease diagnosis. One major challenge for learning-based segmentation method is the limited manually annotated dermoscopy images. Many ...
Comments