# Hi-End-MAE: Hierarchical encoder-driven masked autoencoders are stronger vision learners for medical image segmentation

arXiv: https://arxiv.org/abs/2502.08347 · HTML: https://mlsanity.com/paper/2502.08347/

## Key facts

- Cited by 12 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 12 Feb 2025.

## Built on by (most cited first, 12 of 12)

- [Towards Generalisable Foundation Models for Brain MRI](https://mlsanity.com/paper/2510.23415/) (2025)
- [Pre-Trained LLM is a Semantic-Aware and Generalizable Segmentation Booster](https://mlsanity.com/paper/2506.18034/) (2025)
- [Mobile U-ViT: Revisiting large kernel and U-shaped ViT for efficient medical image segmentation](https://mlsanity.com/paper/2508.01064/) (2025)
- [SimCroP: Radiograph Representation Learning with Similarity-driven Cross-granularity Pre-training](https://mlsanity.com/paper/2509.08311/) (2025)
- [Foundation Model for Whole-Heart Segmentation: Leveraging Student-Teacher Learning in Multi-Modal Medical Imaging](https://mlsanity.com/paper/2503.19005/) (2025)
- [U-RWKV: Lightweight medical image segmentation with direction-adaptive RWKV](https://mlsanity.com/paper/2507.11415/) (2025)
- [MeDUET: Disentangled Unified Pretraining for 3D Medical Image Synthesis and Analysis](https://mlsanity.com/paper/2602.17901/) (2026)
- [ASAP: Advancing Medical Volumetric Representation Learning with Anatomy-aware Semantically-adaptive Pre-training](https://mlsanity.com/paper/2606.00602/) (2026)
- [Vision Foundation Models in Radiology: A Scoping Review of Data, Methodology, Evaluation and Clinical Translation](https://mlsanity.com/paper/2607.07219/) (2026)
- [Decorrelation Speeds Up Vision Transformers](https://mlsanity.com/paper/2510.14657/) (2025)
- [Concept-to-Pixel: Prompt-Free Universal Medical Image Segmentation](https://mlsanity.com/paper/2603.17746/) (2026)
- [Where to Look Matters: Rethinking Sub-Volume Sampling in 3D Medical Self-supervised Learning](https://mlsanity.com/paper/2610.13915/) (2026)

## Abstract (arXiv)

Medical image segmentation remains a formidable challenge due to the label scarcity. Pre-training Vision Transformer (ViT) through masked image modeling (MIM) on large-scale unlabeled medical datasets presents a promising solution, providing both computational efficiency and model generalization for various downstream tasks. However, current ViT-based MIM pre-training frameworks predominantly emphasize local aggregation representations in output layers and fail to exploit the rich representations across different ViT layers that better capture fine-grained semantic information needed for more precise medical downstream tasks. To fill the above gap, we hereby present Hierarchical Encoder-driven MAE (Hi-End-MAE), a simple yet effective ViT-based pre-training solution, which centers on two key innovations: (1) Encoder-driven reconstruction, which encourages the encoder to learn more informative features to guide the reconstruction of masked patches; and (2) Hierarchical dense decoding, which implements a hierarchical decoding structure to capture rich representations across different layers. We pre-train Hi-End-MAE on a large-scale dataset of 10K CT scans and evaluated its performance across seven public medical image segmentation benchmarks. Extensive experiments demonstrate that Hi-End-MAE achieves superior transfer learning capabilities across various downstream tasks, revealing the potential of ViT in medical imaging applications. The code is available at: https://github.com/FengheTan9/Hi-End-MAE