Evaluating the Effectiveness of Graph Neural Networks on an Augmented Dataset for Melanoma Skin Cancer Detection

Benedetto Lozzano; Anvi Kumar; Meet Patel; Arav Kumar; Savya Vats; Avimanyou Vatsa

doi:10.53941/aieng.2026.100007

Abstract

Early detection of malignant melanoma is critical for patient survival, but even trained clinicians achieve only moderate accuracy with traditional visual inspection. Recent work demonstrates that deep convolutional neural networks (CNNs) can rival dermatologists in dermoscopy-based melanoma classification. Nevertheless, CNNs require large, balanced datasets, which are lacking for rare skin lesion classes. This study considered augmenting the ISIC 2019 skin lesion dataset with GAN-generated images to balance the benign and malignant classes and proposed a graph-based classification approach. The dermoscopic images are segmented into super pixels (using Simple Linear Iterative Clustering (SLIC)), a region-adjacency graph of super pixel nodes is constructed, and graph neural networks, including Graph Convolutional Network (GCN), Graph Attention Network (GAT), and a Dynamic Attention Variant graph, are trained. To mitigate the imbalance, we used focal loss and balanced resampling approaches. On the augmented dataset, GNN models achieve higher sensitivity and AUC than a CNN baseline (e.g., GAT ROC = 0.92 vs. CNN ROC = 0.87), indicating the benefit of graph-based relational features. Although based on other experimental results, all models tend to overfit to synthetic images, as performance drops when evaluated on a held-out set of real images. Additionally, training on a smaller, balanced, and real-only subset improves real-world performance and continues to favor GNNs over CNNs. Therefore, these findings suggest that graph representations can enhance melanoma detection in augmented data regimes but also highlight the pitfalls of synthetic augmentation and the need for robust evaluation.

References

1.
CDC. Skin Cancer: Melanoma of the Skin Statistics. Available online: https://www.cdc.gov/skin-cancer/statistics/index.html (accessed on 15 July 2025).
2.
Kumar, A.; Vatsa, A. Untangling Classification Methods for Melanoma Skin Cancer. Front. Big Data 2022, 5, 848614. https://doi.org/10.3389/fdata.2022.848614.
3.
Marghoob, A.A.; Usatine, R.P.; Jaimes, N. Dermoscopy for the Family Physician. Am. Fam. Physician 2013, 88, 441–450.
4.
Godinich, B.M.; Hensperger, V.; Guo, W.; et al. Barriers to Malignant Melanoma Diagnosis in Rural Areas in the United States: A Systematic Review. JAAD Rev. 2024, 1, 29–41.
5.
Petrie, T.; Samatham, R.; Witkowski, A.M.; et al. Melanoma Early Detection: Big Data, Bigger Picture. J. Invest. Dermatol. 2019, 139, 25–30.
6.
Mazhar, T.; Haq, I.; Ditta, A.; et al. The Role of Machine Learning and Deep Learning Approaches for the Detection of Skin Cancer. Healthcare 2023, 11, 415.
7.
Wan, S.; Gong, C.; Zhong, P.; et al. Multi-Scale Dynamic Graph Convolutional Network for Hyperspectral Image Classification. IEEE Trans. Geosci. Remote Sens. 2019, 58, 3162–3177.
8.
Scarselli, F.; Gori, M.; Tsoi, A.C.; et al. The Graph Neural Network Model. IEEE Trans. Neur. Netw. 2009, 20, 61–80. https://doi.org/10.1109/TNN.2008.2005605.
9.
Marino, K.; Salakhutdinov, R.; Gupta, A. The More You Know: Using Knowledge Graphs for Image Classification. In Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu, HI, USA,
21–26 July 2017; pp. 20–28.
10.
Patel, M.; Khurana, D.; Mancinelli, L.; et al. Meta-Analysis on Image Classification Using GNN. https://www.researchsquare.com/article/rs-9817473/v1 (accessed on 16 June 2026).
11.
Joseph, O.; Sukumar, P.; Ulloa, R.; et al. Sentiment Forecasting By Data-Driven Models. In Proceedings of the 35th Wireless and Optical Communications Conference (WOCC), Newark, NJ, USA, 8–9 May 2026.
12.
Codella, N.C.F.; Gutman, D.; Celebi, M.E.; et al. Skin Lesion Analysis Toward Melanoma Detection: A Challenge at the 2017 International Symposium on Biomedical Imaging (ISBI), Hosted by the International Skin Imaging Collaboration (ISIC). In Proceedings of the 2018 IEEE 15th International Symposium on Biomedical Imaging (ISBI 2018), Washington, DC, USA, 4–7 April 2018.
13.
Tschandl, P.; Rosendahl, C.; Kittler, H. The HAM10000 Dataset, a Large Collection of Multi-Source Dermatoscopic Images of Common Pigmented Skin Lesions. Sci. Data 2018, 5, 180161. https://doi.org/10.1038/sdata.2018.161.
14.
Sriram, A.; Vatsa, A.; Kumar, A.; et al. Challenges and Opportunities in Malignant Image Reconstruction Using GAN: A Review. In Proceedings of the 2024 IEEE Integrated STEM Education Conference (ISEC), Princeton, NJ, USA, 9 March 2024; pp. 1–6.
15.
Li, Z.; Chen, Z.; Che, X.; et al. A classification method for multi-class skin damage images combining quantum computing and Inception-ResNet-V1. Front. Phys. 2022, 10. https://doi.org/10.3389/fphy.2022.1046314.
16.
Behara, K.; Bhero, E.; Agee, J.T. Skin Lesion Synthesis and Classification Using an Improved DCGAN Classifier. Diagnostics 2023, 13, 2635.
17.
Skandarani, Y.; Jodoin, P.M.; Lalande, A. GANs for Medical Image Synthesis: An Empirical Study. J. Imaging 2023, 9, 69. https://doi.org/10.3390/jimaging9030069.
18.
Nazir, U.; Wang, H.; Taj, M. Survey of Image Based Graph Neural Networks. arXiv 2021, arXiv:2106.06307.
19.
Shoaib, A.; Vadiveloo, M.; Lim, S.P. Comparative Studies of Region-Based Segmentation Algorithms on Natural and Remote Sensing Images. In Proceedings of the 19th IMT-GT International Conference on Mathematics, Statistics and Their Applications (ICMSA 2024), Bangi, Malaysia, 27–28 May 2024; Volume 67.
20.
Khurana, D.; Patel, M.; Mancinelli, L.; et al. Role of GNN in Skin Cancer Classification. In Proceedings of the 16th IEEE Integrated STEM Education Conference, Princeton, NJ, USA, 14 March 2026.
21.
Khurana, D.; Mittal, H.; Joseph, O.; et al. Empirical Analysis of Video Segmentation for Sustainability. In Proceedings of the 16th IEEE Integrated STEM Education Conference, Princeton, NJ, USA, 14 March 2026.
22.
Gautam, R.R.; Vatsa, A. Meta-Analysis of Image Classification Using Graph Neural Networks (GNNs). In Proceedings of the 16th IEEE Integrated STEM Education Conference, Princeton, NJ, USA, 14 March 2026.
23.
Defferrard, M.; Bresson, X.; Vandergheynst, P. Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering. In Proceedings of the 30th International Conference on Neural Information Processing Systems, Barcelona, Spain, 5–10 December 2016; pp. 3844–3852.
24.
Monroy, L.; Rist, L.; Eberhardt, M.; et al. Employing Graph Representations for Cell-level Characterization of Melanoma MELC Samples. arXiv 2022, arXiv:2211.05884.
25.
ISIC-Archive. ISIC Archive REST API Documentation. Available online: https://www.kaggle.com/c/siim-isic-melanomaclassification/data (accessed on 15 January 2025).
26.
Cinar, U.; Cetin Atalay, R.; Cetin, Y.Y. Human Hepatocellular Carcinoma Classification from H&E Stained Histopathology Images with 3D Convolutional Neural Networks and Focal Loss Function. J. Imaging 2023, 9, 25. https://doi.org/10.3390/jimaging9020025.
27.
Lin, T.Y.; Goyal, P.; Girshick, R.; et al. Focal Loss for Dense Object Detection. IEEE Trans. Pattern Anal. Mach. Intell. 2017, 42, 318–327.
28.
Fey, M.; Lenssen, J.E. Fast Graph Representation Learning with PyTorch Geometric. arXiv 2019, arXiv:1903.02428.
29.
Brody, S.; Alon, U.; Yahav, E. How Attentive Are Graph Attention Networks? arXiv 2022, arXiv:2105.14491.
30.
Sharma, P.; Kumar, M.; Sharma, H.K.; et al. Generative Adversarial Networks (GANs): Introduction, Taxonomy, Variants, Limitations, and Applications. Multimed. Tools Appl. 2024, 83, 88811–88858. https://doi.org/10.1007/s11042-024-18767-y.

Scilight Press

Author Information

Abstract

Keywords

References

About Scilight

Journals

Publishing Policies

Contact Us