![[2505.04586] Active Sampling for MRI-based Sequential Decision Making](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2505.04586] Active Sampling for MRI-based Sequential Decision Making
Summary
This article presents a novel multi-objective reinforcement learning framework for MRI-based sequential decision-making, improving diagnostic capabilities while reducing sampling requirements.
Why It Matters
The advancement of MRI technology as a Point-of-Care (PoC) device is critical for enhancing accessibility and affordability in medical diagnostics. This research addresses the limitations of current MRI practices by proposing a method that optimizes sampling strategies, potentially revolutionizing patient care.
Key Takeaways
- Introduces a reinforcement learning framework for MRI diagnostics.
- Enables sequential decision-making from undersampled k-space data.
- Achieves competitive diagnostic performance with fewer samples.
- Focuses on knee pathology assessment, including ACL and cartilage issues.
- Publicly available code supports further research and application.
Electrical Engineering and Systems Science > Image and Video Processing arXiv:2505.04586 (eess) [Submitted on 7 May 2025 (v1), last revised 13 Feb 2026 (this version, v2)] Title:Active Sampling for MRI-based Sequential Decision Making Authors:Yuning Du, Jingshuai Liu, Rohan Dharmakumar, Sotirios A. Tsaftaris View a PDF of the paper titled Active Sampling for MRI-based Sequential Decision Making, by Yuning Du and 3 other authors View PDF Abstract:Despite the superior diagnostic capability of Magnetic Resonance Imaging (MRI), its use as a Point-of-Care (PoC) device remains limited by high cost and complexity. To enable such a future by reducing the magnetic field strength, one key approach will be to improve sampling strategies. Previous work has shown that it is possible to make diagnostic decisions directly from k-space with fewer samples. Such work shows that single diagnostic decisions can be made, but if we aspire to see MRI as a true PoC, multiple and sequential decisions are necessary while minimizing the number of samples acquired. We present a novel multi-objective reinforcement learning framework enabling comprehensive, sequential, diagnostic evaluation from undersampled k-space data. Our approach during inference actively adapts to sequential decisions to optimally sample. To achieve this, we introduce a training methodology that identifies the samples that contribute the best to each diagnostic objective using a step-wise weighting reward function. We evaluate ou...
