# Attention-Based Chaotic Self-Supervision for Medical Image Classification

arXiv: https://arxiv.org/abs/2605.04985 · HTML: https://mlsanity.com/paper/2605.04985/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 6 of its 12 references that are in the map.
- First posted on arXiv 6 May 2026.

## Builds on (most cited first, 6 of 6)

- [Masked Autoencoders Are Scalable Vision Learners](https://mlsanity.com/paper/2111.06377/) (2021)
- [Learning Imbalanced Datasets with Label-Distribution-Aware Margin Loss](https://mlsanity.com/paper/1906.07413/) (2019)
- [Denoising Diffusion Autoencoders are Unified Self-supervised Learners](https://mlsanity.com/paper/2303.09769/) (2023)
- [ProCo: Prototype-aware Contrastive Learning for Long-tailed Medical Image Classification](https://mlsanity.com/paper/2209.00183/) (2022)
- [Distractor-Aware Neuron Intrinsic Learning for Generic 2D Medical Image Classifications](https://mlsanity.com/paper/2007.09979/) (2020)
- [Fine-Grained Self-Supervised Learning with Jigsaw Puzzles for Medical Image Classification](https://mlsanity.com/paper/2308.05770/) (2023)

## Abstract (arXiv)

Deep learning models for medical image classification usually achieve promising results but typically rely on large, annotated datasets or standard transfer learning from ImageNet. Self-Supervised Learning (SSL) has emerged as a powerful alternative, yet common methods like masked autoencoders (MAEs) may inadvertently destroy fine-grained diagnostic features by using random masking. In this paper, we propose a novel SSL pre-training strategy, the Chaotic Denoising Autoencoder (CDAE). Instead of masking, we apply a chaotic transformation to the input image, tasking an autoencoder to reconstruct the original. We hypothesize this forces the encoder to learn robust, domain-specific features by "inverting the chaos". Furthermore, we propose an attentive fusion mechanism that combines features from our CDAE-trained encoder with a standard encoder, leveraging the strengths of both general and domain-specific representations. Our method is evaluated on two public medical datasets: ISIC 2018 (skin lesions) and APTOS 2019 (diabetic retinopathy). The proposed model achieves high performance, with an accuracy of 0.9221 and an F1-macro of 0.8530 on ISIC 2018, and an accuracy of 0.8644 and F1-macro of 0.7433 on APTOS 2019, demonstrating the efficacy of our approach.