Anatomical Token Uncertainty for Transformer-Guided Active MRI Acquisition

2026-03-23

AI Summary

This work addresses the prolonged acquisition time of fully sampled magnetic resonance imaging (MRI), which compromises clinical efficiency and patient comfort. To mitigate this, we propose a novel active sampling framework leveraging a pretrained medical image tokenizer and a latent-space Transformer. For the first time, visual token entropy derived from anatomical structures is employed as an uncertainty metric to guide adaptive k-space sampling. We introduce two complementary strategies—Latent Entropy Selection and Gradient-based Entropy Optimization—to maximize the information content of acquired samples. Evaluated on the fastMRI knee and brain datasets, our method consistently outperforms state-of-the-art approaches at both 8× and 16× acceleration rates, achieving significant improvements in perceptual quality and structural fidelity.

Technology Category

• Machine Learning: Active Learning
• Search and Optimization: Sampling/Simulation-based Search
• Computer Vision: Medical and Biological Imaging

Application Category

• Search and Retrieval-Augmented AI: Agentic search
• User Modeling, Personalization and Recommendation: Accountability, Transparency, and Ethics for personalization
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Full data acquisition in MRI is inherently slow, which limits clinical throughput and increases patient discomfort. Compressed Sensing MRI (CS-MRI) seeks to accelerate acquisition by reconstructing images from under-sampled k-space data, requiring both an optimal sampling trajectory and a high-fidelity reconstruction model. In this work, we propose a novel active sampling framework that leverages the inherent discrete structure of a pretrained medical image tokenizer and a latent transformer. By representing anatomy through a dictionary of quantized visual tokens, the model provides a well-defined probability distribution over the latent space. We utilize this distribution to derive a principled uncertainty measure via token entropy, which guides the active sampling process. We introduce two strategies to exploit this latent uncertainty: (1) Latent Entropy Selection (LES), projecting patch-wise token entropy into the $k$-space domain to identify informative sampling lines, and (2) Gradient-based Entropy Optimization (GEO), which identifies regions of maximum uncertainty reduction via the $k$-space gradient of a total latent entropy loss. We evaluate our framework on the fastMRI singlecoil Knee and Brain datasets at $\times 8$ and $\times 16$ acceleration. Our results demonstrate that our active policies outperform state-of-the-art baselines in perceptual metrics, and feature-based distances. Our code is available at https://github.com/levayz/TRUST-MRI.

Problem

• MRI acceleration
• compressed sensing
• active sampling
• k-space undersampling
• image reconstruction

Innovation

• Active MRI Sampling
• Token Entropy
• Latent Transformer
• Compressed Sensing MRI
• Uncertainty-Guided Acquisition