An Evaluation of Transfer Learning With CheXNet on Lung Opacity Detection in COVID-19 and Pneumonia Chest Radiographs
Bibliographic record
Abstract
As the COVID-19 pandemic continues to put immense stress on hospitals, healthcare workers, and intensive care units, a quick diagnosis and disease severity assessment for patients is crucial. This would allow clinicians to provide the right treatment early on, and thus, prevent serious illness in patients later. Chest radiography is a fast method of diagnosing patients. By analyzing the presence and distribution of lung opacities in chest radiographs, clinicians can determine the severity of COVID-19 or other pneumonia and apply proper treatment early on. Hence, research into models that can detect such lung opacities in chest radiographs would help clinicians efficiently diagnose. Currently, much research is being done on gathering and classifying chest radiographs. A milestone in this regard has been the development of CheXNet by the Stanford ML group, which claims better performance than radiologists at classifying chest radiographs. In this study, the CheXNet feature extractor backbone is used to test if it can improve the performance of lung opacity object detection models with transfer learning. No improvement in performance was observed on a variety of test datasets, with the models trained using the CheXNet feature extractor experiencing a slight decrease in performance on some test datasets.
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 | 0.007 | 0.012 |
| Meta-epidemiology (narrow) | 0.003 | 0.000 |
| Meta-epidemiology (broad) | 0.002 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.003 | 0.002 |
| Insufficient payload (model declined to judge) | 0.003 | 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".