Skip to content

3 papers indexed here

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

Aug 2026

Dual-Echo bSSFP for Rapid T2-Sensitive Quantitative Contrast Imaging at Ultra-Low Field.

OBJECTIVE Rapid characterization of transverse-relaxation-sensitive MRI contrast is important for evaluating tissue-dependent signal behavior, but remains challenging in ultra-low-field (ULF) MRI because of limited signal-to-noise ratio (SNR) and acquisition-efficiency constraints. This study aims to develop a rapid, high-SNR, sequence-specific T2-sensitive quantitative contrast imaging method for ULF MRI. METHODS A balanced dual-echo steady-state (bDESS) sequence was developed to acquire two echoes at predefined echo times within each repetition of a balanced steady-state free precession acquisition. A logarithmic-ratio operator, based on a mono-exponential attenuation approximation, was used to derive a sequence-specific T2-sensitive contrast index, termed T2SDI, from the two echo magnitudes. The proposed method was implemented on a custom-built 6.5 mT MRI system and evaluated using numerical simulations, CuSO₄ phantom experiments, and in vivo brain imaging. RESULTS Numerical simulations showed that T2SDI exhibited a monotonic and approximately linear dependence on T2 under controlled field-inhomogeneity conditions. Phantom experiments confirmed that bDESS-derived T2SDI increased with CPMG-measured reference T2 and showed higher SNR than dual-echo SPGR. The method was further demonstrated in vivo by generating T2SDI maps of the human brain. CONCLUSION This study presents a sequence-specific, index-based method for rapid T2-sensitive quantitative contrast imaging in ULF MRI. SIGNIFICANCE The proposed dual-echo bSSFP/bDESS method provides a high-SNR and time-efficient strategy for sequence-specific T2-sensitive contrast characterization at ultra-low field.

Sheng Shen, Neha Koonjoo, Hester A. Braaksma et al. · 0 citations
Open access Jul 2026

Addressing benchmarking gaps in large language models for health and medicine with dynamic red-teaming

A Dynamic, Automatic and Systematic red-teaming audit framework that continuously stress-tests LLMs for health across four safety-critical axes: robustness, privacy, bias and hallucination, which provides a scalable framework for surfacing latent risks before such systems are deployed in consumer-facing health assistants and broader clinical workflows.

Jiazhen Pan, Bailiang Jian, Paul Hager et al. · 0 citations
Preprint Jul 2026

Med-OPD: Improving Medical Vision-Language Models via Evidence-Aware On-Policy Distillation

Medical Vision-Language Models (Med-VLMs) require reliable reasoning from fine-grained visual evidence, yet existing models can produce plausible clinical answers by relying on language priors or medical templates rather than truly attending to diagnosis-critical regions. On-Policy Distillation (OPD) offers dense token-level supervision on student-generated trajectories and provides a privacy-compatible means of capability transfer without requiring the redistribution of raw patient data. However, standard OPD uniformly distills all tokens, causing sparse evidence-dependent tokens to be diluted by abundant clinical narrative tokens. Inspired by the success of OPD in the large language model community, we propose \textbf{Med-OPD}, to our knowledge the first unified post-training framework that integrates on-policy distillation with medical evidence-aware supervision for Med-VLMs. We introduce \textbf{Medical Evidence Advantage} (MEA), a teacher-grounded counterfactual signal that uses an answer-aware hint to focus teacher scoring on evidence supporting the target diagnosis, and measures each token's dependence on medical visual evidence by comparing teacher likelihoods under the original and evidence-degraded imaging modalities. Based on MEA, Med-OPD redistributes the distillation signal at both the token and trajectory levels, emphasizing diagnosis-critical tokens and evidence-reliant rollouts. Experiments on OmniMedVQA subsets show that Med-OPD consistently outperforms SFT and standard OPD across CT, MRI, Disease Diagnosis, and Lesion Grading. These results demonstrate that evidence-aware distillation can better strengthen medical VLMs'reliance on key visual evidence and improve reliable multimodal medical reasoning. The source code and data is publicly available at: https://github.com/yunhang8658/MedOPD.git

Yunhang Qian, Jiaquan Yu, Jiawei Liu et al. · 1 citation