# Weight Group-wise Post-Training Quantization for Medical Foundation Model

arXiv: https://arxiv.org/abs/2604.07674 · HTML: https://mlsanity.com/paper/2604.07674/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 24 of its 30 references that are in the map.
- First posted on arXiv 9 Apr 2026.

## Builds on (most cited first, 24 of 24)

- [Segment Anything](https://mlsanity.com/paper/2304.02643/) (2023)
- [GPTQ: Accurate Post-Training Quantization for Generative Pre-trained Transformers](https://mlsanity.com/paper/2210.17323/) (2022)
- [AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration](https://mlsanity.com/paper/2306.00978/) (2023)
- [Quantizing deep convolutional networks for efficient inference: A whitepaper](https://mlsanity.com/paper/1806.08342/) (2018)
- [A White Paper on Neural Network Quantization](https://mlsanity.com/paper/2106.08295/) (2021)
- [Fast Segment Anything](https://mlsanity.com/paper/2306.12156/) (2023)
- [AbdomenCT-1K: Is Abdominal Organ Segmentation A Solved Problem?](https://mlsanity.com/paper/2010.14808/) (2020)
- [Efficient Deep Learning: A Survey on Making Deep Learning Models Smaller, Faster, and Better](https://mlsanity.com/paper/2106.08962/) (2021)
- [Deep Semantic Segmentation of Natural and Medical Images: A Review](https://mlsanity.com/paper/1910.07655/) (2019)
- [SAM2-Adapter: Evaluating & Adapting Segment Anything 2 in Downstream Tasks: Camouflage, Shadow, Medical Image Segmentation, and More](https://mlsanity.com/paper/2408.04579/) (2024)
- [MobileSAMv2: Faster Segment Anything to Everything](https://mlsanity.com/paper/2312.09579/) (2023)
- [A Comprehensive Study on Quantization Techniques for Large Language Models](https://mlsanity.com/paper/2411.02530/) (2024)
- [MagR: Weight Magnitude Reduction for Enhancing Post-Training Quantization](https://mlsanity.com/paper/2406.00800/) (2024)
- [Model Compression in Practice: Lessons Learned from Practitioners Creating On-device Machine Learning Experiences](https://mlsanity.com/paper/2310.04621/) (2023)
- [EasyQuant: An Efficient Data-free Quantization Algorithm for LLMs](https://mlsanity.com/paper/2403.02775/) (2024)
- [UU-Mamba: Uncertainty-aware U-Mamba for Cardiac Image Segmentation](https://mlsanity.com/paper/2405.17496/) (2024)
- [COMQ: A Backpropagation-Free Algorithm for Post-Training Quantization](https://mlsanity.com/paper/2403.07134/) (2024)
- [Robust COVID-19 Detection in CT Images with CLIP](https://mlsanity.com/paper/2403.08947/) (2024)
- [Robustly Optimized Deep Feature Decoupling Network for Fatty Liver Diseases Detection](https://mlsanity.com/paper/2406.17338/) (2024)
- [UMedNeRF: Uncertainty-aware Single View Volumetric Rendering for Medical Neural Radiance Fields](https://mlsanity.com/paper/2311.05836/) (2023)
- [Diffusion-empowered AutoPrompt MedSAM](https://mlsanity.com/paper/2502.06817/) (2025)
- [Neural Radiance Fields in Medical Imaging: A Survey](https://mlsanity.com/paper/2402.17797/) (2024)
- [Improving Generalization of Medical Image Registration Foundation Model](https://mlsanity.com/paper/2505.06527/) (2025)
- [Is (Selective) Round-To-Nearest Quantization All You Need?](https://mlsanity.com/paper/2505.15909/) (2025)

## Abstract (arXiv)

Foundation models have achieved remarkable results in medical image analysis. However, its large network architecture and high computational complexity significantly impact inference speed, limiting its application on terminal medical devices. Quantization, a technique that compresses models into low-bit versions, is a solution to this challenge. In this paper, we propose a post-training quantization algorithm, Permutation-COMQ. It eliminates the need for backpropagation by using simple dot products and rounding operations, thereby removing hyperparameter tuning and simplifying the process. Additionally, we introduce a weight-aware strategy that reorders the weight within each layer to address the accuracy degradation induced by channel-wise scaling during quantization, while preserving channel structure. Experiments demonstrate that our method achieves the best results in 2-bit, 4-bit, and 8-bit quantization.