Toward Vision Language Model-based Assessment of Clinical Quality and Usability of LGE-MR Images for Cardiac Ablation Planning

Bipasha Kundu; Abhishek Chaturvedi; Axel W. E. Wismueller; Richard Simon; Cristian A. Linte

eess.IVAug 21, 2026

Toward Vision Language Model-based Assessment of Clinical Quality and Usability of LGE-MR Images for Cardiac Ablation Planning

Authors: Bipasha Kundu, Abhishek Chaturvedi, Axel W. E. Wismueller, Richard Simon, Cristian A. Linte

Organizations: Carlson Centre of Imaging Sceince, RIT, Rochester, NY, USA · Department of Imaging Sciences, University of Rochester Medical Center, NY, USA · Biomedical Engineering, RIT, Rochester, NY, USA

Abstract

LGE cardiac MRI is widely used for left atrial fibrosis assessment and ablation planning in atrial fibrillation patients as knowledge of fibrotic tissue regions identified from LGE-MRI is critical for catheter ablation. Often, poor quality images used during ablation planning can cause mis-localization of ablation targets, directly impacting procedure safety and outcome. The decision of whether a scan meets the minimum quality threshold for ablation planning is currently made informally by the reviewing radiologist and is not captured by any automated system, yet it is arguably the most safety-critical output of the image quality assessment (IQA) process. However, variations in image quality caused by noise, motion artifacts, and poor boundary definition significantly compromise the reliability of downstream segmentation and clinical decision-making tasks. Manual quality assessment by expert radiologists is subjective and difficult to scale, while existing automated methods produce scalar scores without interpretable clinical reasoning. In this work, we propose a two-stage vision language model (VLM) framework for clinically grounded image quality assessment of left atrial LGE-MRI. In the first stage, a fine-tuned VLM generates structured radiology-style quality reports predicting five radiologist-defined criteria: Noise, Motion Artifact, LA Boundary Accuracy, PV Region Accuracy, and Under-segmentation Severity. In the second stage, a GPT-based reasoning module maps the predicted quality and reports to a structured quality scores and binary clinical usability decision for ablation planning. We curate a dataset of 60 annotated image slice-text pairs from 20 patients and benchmark four state-of-the-art VLM architectures. InternVL2 achieves the highest criterion-level accuracy (Avg ACC=0.65, PLCC=0.79), while DeepSeek achieves perfect clinical usability agreement (Acc=1.00, kappa=1.00).

Figures & tables

Criterion	Meaning	Score Range
Noise	Level of graininess or signal disturbance affecting visibility of the left atrial wall	0–3
Motion Artifact	Presence of motion-induced blurring or misalignment affecting the left atrial wall	0–3
LA Boundary
Accuracy	Clarity and continuity of the left atrial wall boundary	0–3

PV Region

Table 1: Definition of quality assessment criteria used in this study.

Figure 1: Distribution of the radiologist-assigned IQA scores across the five criteria during the construction of the MRI-IQA dataset.

Figure 2: Overview of IQA pipeline. Data curation and annotation are followed by LoRA-adapted fine-tuning of the VLM that generates scores and textual reports, which are processed by an LLM to produce a clinical usability decision.

Model	Metric	Noise	Motion	Boundary	PV Region	Under-seg	Avg
InternVL2-2B	ACC $\uparrow$	0.75	0.75	0.67	0.50	0.58	0.65
	PLCC $\uparrow$	0.91	0.91	0.75	0.71	0.72	0.79
	Tol ${}_{\text{Acc}}$ $\uparrow$	1.00	1.00	1.00	1.00	1.00	1.00
DeepSeek-v-1.3b	ACC $\uparrow$	0.75	0.75	0.50	0.50	0.50	0.60
	PLCC $\uparrow$	0.91	0.91	0.71	0.71	0.71	0.78
	Tol ${}_{\text{Acc}}$ $\uparrow$	1.00	1.00	1.00	1.00	1.00	1.00

Table 2: Performance across models for each criterion (Noise, Motion, LA Boundary Accuracy, PV Region Visibility, and Under-segmentation Severity) assessed using Accuracy [ACC

\uparrow

], Pearson Linear Correlation Coefficient [PLCC

\uparrow

], and Tolerance Accuracy [Tol Acc

\uparrow

], where

\uparrow

indicates higher is better.

Method	Acc $\uparrow$	F1 $\uparrow$	Cohen’s
		(Y/N)	Kappa $\uparrow$
DeepSeek-v-1.3b	1.00	1.00	1.00
InternVL2-2B	0.92	0.90	0.83
LLaVa-1.5-7b	0.75	0.80	0.50
Qwen2.5-3B	0.83	0.80	0.67

Table 3: Final clinical usability decision performance across methods.

Figure 3: Qualitative comparison of VLM Prediction, Chat-GPT extracted scores and clinical decision for usability with ground truth

Explore similar work

Apr 11, 2026

AC-MIL: Weakly Supervised Atrial LGE-MRI Quality Assessment via Adversarial Concept Disentanglement

High-quality Late Gadolinium Enhancement (LGE) MRI can be helpful for atrial fibrillation management, yet scan quality is frequently compromised by patient motion, irregular breathing, and suboptimal image acquisition timing. While Multiple Instance Learning (MIL) has emerged as a powerful tool for automated quality assessment under weak supervision, current state-of-the-art methods map localized visual evidence to a single, opaque global feature vector. This black box approach fails to provide actionable feedback on specific failure modes, obscuring whether a scan degrades due to motion blur, inadequate contrast, or a lack of anatomical context. In this paper, we propose Adversarial Concept-MIL (AC-MIL), a weakly supervised framework that decomposes global image quality into clinically defined radiological concepts using only volume-level supervision. To capture latent quality variations without entangling predefined concepts, our framework incorporates an unsupervised residual branch guided by an adversarial erasure mechanism to strictly prevent information leakage. Furthermore, we introduce a spatial diversity constraint that penalizes overlap between distinct concept attention maps, ensuring localized and interpretable feature extraction. Extensive experiments on a clinical dataset of atrial LGE-MRI volumes demonstrate that AC-MIL successfully opens the MIL black box, providing highly localized spatial concept maps that allow clinicians to pinpoint the specific causes of non-diagnostic scans. Crucially, our framework achieves this deep clinical transparency while maintaining highly competitive ordinal grading performance against existing baselines. Code to be released on acceptance.

K M Arefeen Sultan, Kaysen Hansen, Benjamin Orkild +7

Scientific Computing and Imaging Institute, University of Utah, SLC, UT · Kahlert School of Computing, University of Utah, SLC, UT · Department of Biomedical Engineering, University of Utah, SLC, UT +4

Medical Image Analysis Explainable Medical Image Analysis Multiple Instance Learning Cardiac MRI Weakly Supervised Learning

Ask PDF

Oct 1, 2026

CineMR: Tool-Integrated Vision-Language Reasoning for Quantitative Cardiac MRI Assessment

Cardiovascular magnetic resonance (CMR), including cine imaging, is a reference standard for the noninvasive assessment of cardiac morphology and ventricular function. Cine CMR interpretation integrates qualitative visual assessment with quantitative measurements of ventricular volumes, ejection fraction, myocardial mass, wall thickness, and regional wall motion. Current medical vision-language models (VLMs) cannot reliably derive quantitative measurements from multidimensional cine images without analysis tools. We present CineMR, a tool-augmented VLM that invokes cardiac image-analysis tools and integrates their outputs into interleaved reasoning for quantitative CMR assessment. We also construct a multi-cohort visual question answering benchmark covering quantitative metric extraction, multiclass diagnosis, and differential diagnosis, together with tools for segmentation, phase selection, volumetry, morphometry, and regional wall motion analysis. CineMR is trained with supervised fine-tuning (SFT) on tool-interaction traces followed by Group Relative Policy Optimization (GRPO) with conditional tool-use rewards. On the multi-cohort cine CMR benchmark, CineMR achieves 35.9% pass@1 and 58.9% pass@4, compared with 1.5% pass@1 for the Qwen3-VL-8B backbone and 0.0% and 7.0% pass@1 for LLaVA-Med v1.5 and MedGemma-4B, respectively. Correct tool invocation reaches 99.8% after GRPO, up from 78.9% after SFT. Live tool outputs improve ventricular measurement accuracy by 20.4--23.7% over direct model predictions, and removing all tools reduces pass@1 from 35.9% to 27.9%. These results highlight the importance of reliable tool use for quantitative cine CMR reasoning and support CineMR as a promising approach for assistive cardiac image assessment. Code, benchmark resources, and model weights are available at https://github.com/AI-MIND-Lab/CineMR.

Kunyang Li, Hai Nguyen, Joshua Lowe +6

Institute for Artificial Intelligence, University of Central Florida · Department of Clinical Sciences, College of Medicine, University of Central Florida · Nemours Cardiac Center, Nemours Children’s Hospital, Florida +1

Tool Use in VLMs Medical VQA VLM Reasoning Medical Image Benchmarks Cardiac MRI

Ask PDF

Sep 18, 2026

Multicentre Bi-atrial Segmentation from LGE-MRI for Atrial Fibrillation with a 2D and 3D Framework

Accurate delineation of bi-atrial structures from late gadolinium enhancement MRI (LGE-MRI) is an important prerequisite for structural analysis and future fibrosis-quantification workflows in atrial fibrillation (AF). However, automated segmentation is challenging due to thin-walled anatomy, domain shifts across imaging centres, and limited benchmarking of existing methods. This study presents a two-stage segmentation framework and benchmarking platform for evaluating how ROI localisation, encoder design, 2D/3D dimensionality, and ensemble fusion affect bi-atrial wall and cavity segmentation across multicentre LGE-MRI datasets. The framework integrates 3D localisation and fine segmentation using 2D and 3D U-Net variants with ResNeXt encoders and compares them with convolutional, transformer-based, and state-space architectures. { Evaluation across three independent cohorts assessed accuracy and cross-domain transfer without target-domain fine-tuning. Cavity segmentation transferred more consistently across centres than atrial wall segmentation, while wall performance remained sensitive to domain shift, particularly in the Kobe cohort.} By quantifying how 2D, 3D, and ensemble architectures behave across centres and between walls and cavities, this work provides a reproducible benchmark for future methodological development and clinical validation.

Malitha Gunawardhana, Gregory B. Sands, Mark L. Trew +1

Medical Image Segmentation Medical Image Benchmarks Cardiac MRI

Ask PDF