Research map / Image generation / Text-to-image diffusion
Learning to Read Chest X-Rays: Recurrent Neural Cascade Model for Automated Image Annotation
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use joint image/text context vectors to relabel images and retrain the CNN/RNN | Chest X-ray annotation generation with GRU · Chest X-ray test | BLEU-2: 3.0 -> 14.4 | Helped |
| Use joint image/text context vectors to relabel images and retrain the CNN/RNN | Chest X-ray annotation generation with GRU · Chest X-ray test | BLEU-3: 0.0 -> 4.7 | Helped |
| Use joint image/text context vectors to relabel images and retrain the CNN/RNN | Chest X-ray annotation generation with GRU · Chest X-ray test | BLEU-1: 75.4 -> 78.5 | Helped |
| Use joint image/text context vectors to relabel images and retrain the CNN/RNN | Chest X-ray annotation generation with LSTM · Chest X-ray test | BLEU-2: 0.4 -> 9.1 | Helped |
| Use joint image/text context vectors to relabel images and retrain the CNN/RNN | Chest X-ray annotation generation with LSTM · Chest X-ray test | BLEU-1: 78.3 -> 79.3 | Helped |
| Combine batch normalization and data-dropout (ablation) | NIN disease classifier · Chest X-ray validation | accuracy: 58.38% -> 62.43% | Helped |
| Add batch normalization to data-dropout (ablation) | NIN disease classifier · Chest X-ray validation | accuracy: 58.38% -> 62.43% | Helped |
| Add data-dropout to batch normalization (ablation) | NIN disease classifier · Chest X-ray validation | accuracy: 56.65% -> 62.43% | Helped |
| Use GoogLeNet instead of NIN with BN and data-dropout (ablation) | Chest X-ray disease classifier · Chest X-ray validation | accuracy: 62.43% -> 66.40% | Helped |
| Duplicate rather than crop training images for balancing (ablation) | GoogLeNet disease classifier with BN and data-dropout · Chest X-ray validation | accuracy: 66.40% -> 69.84% | Helped |
| Add dropout to the LSTM (ablation) | LSTM annotation generator · Annotation generation validation | validation loss | Hurt |
| Increase the LSTM to three layers (ablation) | LSTM annotation generator · Annotation generation validation | validation loss | Hurt |
Closest papers by meaning
- Bootstrapping Chest CT Image Understanding by Distilling Knowledge from X-ray Expert Models (CVPR 2024 · similarity 0.88)
- Learning Low-Rank Feature for Thorax Disease Classification (NeurIPS 2024 · similarity 0.86)
- Novel Class Discovery in Chest X-rays via Paired Images and Text (AAAI 2024 · similarity 0.86)
- Large-scale and Fine-grained Vision-language Pre-training for Enhanced CT Image Understanding (ICLR 2025 · similarity 0.85)
- RadZero: Similarity-Based Cross-Attention for Explainable Vision-Language Alignment in Chest X-ray with Zero-Shot Multi-Task Capability (NeurIPS 2025 · similarity 0.85)
- HC-LLM: Historical-Constrained Large Language Models for Radiology Report Generation (AAAI 2025 · similarity 0.85)
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.