A CNN-Based Framework for Automatic Segmentation of Chest X-ray Images and Multi-Type Respiratory Disease Recognition
Bibliographic record
Abstract
Respiratory diseases pose a significant threat to human health, and early, accurate diagnosis is critical for improving patient outcomes.Chest X-ray imaging, due to its affordability and convenience, remains the primary tool for clinical screening.However, traditional manual interpretation is time-consuming, experience-dependent, and prone to oversight and misdiagnosis.With the advancement of deep learning in medical imaging, computer visionbased automated analysis offers promising solutions to these challenges.Nevertheless, existing methods such as U-Net and its variants often struggle with accurately segmenting complex lung structures, especially when dealing with noisy images, small lesions, or blurred boundaries.Additionally, conventional 2D convolutional neural networks (CNNs) have limitations in capturing the spatial features inherent in chest X-ray images, and current multi-disease classification models still face challenges in achieving high accuracy and generalizability.To address these issues, this study proposes two key innovations: First, an optimized U-Net++L3 network with pruning is developed for automatic chest X-ray segmentation, effectively reducing parameter redundancy while maintaining accuracy, thereby enhancing segmentation performance in regions with complex lesions.Second, a densely connected 3D CNN model is designed for the recognition of multiple respiratory diseases.By leveraging the spatial feature extraction capabilities of 3D convolutions and the feature reuse advantages of dense connections, the model achieves precise classification of conditions such as pneumonia, lung cancer, and chronic obstructive pulmonary disease (COPD).The outcomes of this research aim to overcome the limitations of traditional models in terms of segmentation accuracy, computational efficiency, and feature representation, providing both theoretical innovation and practical value for rapid clinical screening and the enhancement of primary healthcare resources.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".