# Geometry-Aware Superpixel Graph Transformer with Metadata for Skin Lesion Classification

arXiv: https://arxiv.org/abs/2606.20390 · HTML: https://mlsanity.com/paper/2606.20390/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 9 of its 36 references that are in the map.
- First posted on arXiv 18 Jun 2026.

## Builds on (most cited first, 9 of 9)

- [Semi-Supervised Classification with Graph Convolutional Networks](https://mlsanity.com/paper/1609.02907/) (2016)
- [Masked Label Prediction: Unified Message Passing Model for Semi-Supervised Classification](https://mlsanity.com/paper/2009.03509/) (2020)
- [PAD-UFES-20: a skin lesion dataset composed of patient data and clinical images collected from smartphones](https://mlsanity.com/paper/2007.00478/) (2020)
- [Attention-based Graph Neural Network for Semi-supervised Learning](https://mlsanity.com/paper/1803.03735/) (2018)
- [Soft-Attention Improves Skin Cancer Classification Performance](https://mlsanity.com/paper/2105.03358/) (2021)
- [Deeply Supervised Skin Lesions Diagnosis with Stage and Branch Attention](https://mlsanity.com/paper/2205.04326/) (2022)
- [Key Patches Are All You Need: A Multiple Instance Learning Framework For Robust Medical Diagnosis](https://mlsanity.com/paper/2405.01654/) (2024)
- [Diffusion models applied to skin and oral cancer classification](https://mlsanity.com/paper/2504.00026/) (2025)
- [Skin Lesion Phenotyping via Nested Multi-modal Contrastive Learning](https://mlsanity.com/paper/2505.23709/) (2025)

## Abstract (arXiv)

Automated skin cancer classification from dermoscopic images remains challenging due to heterogeneous lesion structure, strong intra-class variability, and subtle visual differences between benign and malignant cases. Existing CNN/ViT pipelines typically rely on global or patch-level features and often combine patient metadata via late fusion, which limits spatially grounded multimodal reasoning. We present a novel region-based graph learning framework that explicitly models lesions as graphs of spatially coherent superpixel regions represented as frozen CNN features. To capture fine-grained lesion arrangements, we encode inter-regional geometry as edge attributes and introduce a dedicated metadata context node connected to all regions, providing structured integration of demographic/clinical variables within the same relational space. Node representations are updated using our edge-aware graph transformer followed by attention-driven propagation, and a final graph-level embedding for benign-malignant classification. Experiments on four public benchmarks demonstrate that explicit region-level relational modeling and graph-native multimodal fusion yield consistent gains over the state-of-the-art. Consequently, we establish a new graph-centric perspective in which CNN features are modeled as relational nodes and improved through contextual integration, yielding more expressive and robust classifications.