AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

RefleXNet: Targeted Self-Reflection for Accurate Chest X-ray Reporting

Xin Mei, Rui Mao, Xiaoyan Cai, Libin Yang, Erik Cambria

AAAI 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Multimodal LLM benchmarks · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes RefleXNet, a chest X-ray report generation model with targeted self-reflection to better capture subtle abnormalities, as stated in the truncated abstract.

Problem

Automated chest X-ray report generation relies on global visual features and token-level supervision, which limits sensitivity to subtle abnormalities and clinical reliability.

Approach

Introduces targeted self-reflection for CXR reporting as a way to move beyond global visual features and token-level supervision; specific mechanism not stated in the available abstract.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add targeted self-reflection optimization after SFT (ablation)RefleXNet 3B · MIMIC-CXRF1-RadGraph: 0.291 -> 0.311Improved
add targeted self-reflection optimization (ablation)RefleXNet 3B · MIMIC-CXRCheXbert micro-F1 (14 obs.): 0.505 -> 0.570Improved
add targeted self-reflection optimization (ablation)RefleXNet 3B · MIMIC-CXRRadCliQ-v0: 3.15 -> 2.96Improved
add targeted self-reflection optimization (ablation)RefleXNet 3B · MIMIC-CXR pneumothoraxF1: 0.1696 -> 0.3385Improved
add targeted self-reflection optimization (ablation)RefleXNet 3B · MIMIC-CXR cardiomegalyF1: 0.6433 -> 0.6772Improved
replace BaML loss with standard BCE (ablation)MGVE · MIMIC-CXR abnormality predictionmacro-F1: 0.388 -> 0.264Hurt
remove patch-level feature fusion (region-only) (ablation)MGVE · MIMIC-CXR abnormality predictionmicro-F1: 0.533 -> 0.476Hurt
add region-abnormality mapping matrix (ablation)MGVE · MIMIC-CXR abnormality predictionmacro-F1: 0.357 -> 0.388Improved
remove mapping matrix (ablation)MGVE · MIMIC-CXR abnormality predictionmicro-precision: 0.463 -> 0.501Mixed
add two extra epochs of self-reflection on IU-XrayRefleXNet · IU-XrayF1-RadGraph: 0.325 -> 0.384Improved
add self-reflection to the zero-shot model (ablation)RefleXNet on IU-Xray · IU-Xray zero-shotCheXbert vector similarity: 0.590 -> 0.585No clear effect
continue training Qwen2.5-VL on MIMIC-CXRQwen2.5-VL 3B · MIMIC-CXRF1-RadGraph: 0.094 -> 0.263Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy