GDMRG integrates a Topological Knowledge Internalization module using GCN on disease co-occurrence priors with dual-stream classifiers and diagnosis-guided spatial attention to produce more accurate and fluent medical reports from chest X-rays, with competitive results on MIMIC-CXR and zero-shot on
MIMIC-CXR, a de- identified publicly available database of chest radio- graphs with free-text reports
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Graph-Augmented Topological Internalization with Dual-Stream Classifiers for Medical Report Generation
GDMRG integrates a Topological Knowledge Internalization module using GCN on disease co-occurrence priors with dual-stream classifiers and diagnosis-guided spatial attention to produce more accurate and fluent medical reports from chest X-rays, with competitive results on MIMIC-CXR and zero-shot on