# Memorizing SAM: 3D Medical Segment Anything Model with Memorizing Transformer

arXiv: https://arxiv.org/abs/2412.13908 · HTML: https://mlsanity.com/paper/2412.13908/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 13 of its 19 references that are in the map.
- First posted on arXiv 18 Dec 2024.

## Builds on (most cited first, 13 of 13)

- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [U-Net: Convolutional Networks for Biomedical Image Segmentation](https://mlsanity.com/paper/1505.04597/) (2015)
- [Segment Anything](https://mlsanity.com/paper/2304.02643/) (2023)
- [TotalSegmentator: robust segmentation of 104 anatomical structures in CT images](https://mlsanity.com/paper/2208.05868/) (2022)
- [Segment Anything Model for Medical Image Analysis: an Experimental Study](https://mlsanity.com/paper/2304.10517/) (2023)
- [Memorizing Transformers](https://mlsanity.com/paper/2203.08913/) (2022)
- [MedNeXt: Transformer-driven Scaling of ConvNets for Medical Image Segmentation](https://mlsanity.com/paper/2303.09975/) (2023)
- [SAM-Med2D](https://mlsanity.com/paper/2308.16184/) (2023)
- [SAM3D: Segment Anything Model in Volumetric Medical Images](https://mlsanity.com/paper/2309.03493/) (2023)
- [A Survey on Deep Learning for Skin Lesion Segmentation](https://mlsanity.com/paper/2206.00356/) (2022)
- [TransNuSeg: A Lightweight Multi-Task Transformer for Nuclei Segmentation](https://mlsanity.com/paper/2307.08051/) (2023)
- [FastSAM3D: An Efficient Segment Anything Model for 3D Volumetric Medical Images](https://mlsanity.com/paper/2403.09827/) (2024)
- [MoViT: Memorizing Vision Transformers for Medical Image Analysis](https://mlsanity.com/paper/2303.15553/) (2023)

## Abstract (arXiv)

Segment Anything Models (SAMs) have gained increasing attention in medical image analysis due to their zero-shot generalization capability in segmenting objects of unseen classes and domains when provided with appropriate user prompts. Addressing this performance gap is important to fully leverage the pre-trained weights of SAMs, particularly in the domain of volumetric medical image segmentation, where accuracy is important but well-annotated 3D medical data for fine-tuning is limited. In this work, we investigate whether introducing the memory mechanism as a plug-in, specifically the ability to memorize and recall internal representations of past inputs, can improve the performance of SAM with limited computation cost. To this end, we propose Memorizing SAM, a novel 3D SAM architecture incorporating a memory Transformer as a plug-in. Unlike conventional memorizing Transformers that save the internal representation during training or inference, our Memorizing SAM utilizes existing highly accurate internal representation as the memory source to ensure the quality of memory. We evaluate the performance of Memorizing SAM in 33 categories from the TotalSegmentator dataset, which indicates that Memorizing SAM can outperform state-of-the-art 3D SAM variant i.e., FastSAM3D with an average Dice increase of 11.36% at the cost of only 4.38 millisecond increase in inference time. The source code is publicly available at https://github.com/swedfr/memorizingSAM