CT vessel features improves COPD classification
Bibliographic record
Abstract
Background: Computed tomography (CT) features extracted from the lung and airways can classify COPD with machine learning (ML) models (Puchakayala Radiol 2023). However, pulmonary vasculature changes occur in early COPD (Santos Eur Respir J 2002), and it is unknown if CT pulmonary vessel features can improve COPD classification. Aim: To compare ML models for classifying COPD using combinations of CT features extracted from the lung, airways, and vessels. We hypothesize that the inclusion of CT vessel features to lung and/or airway features will improve ML performance for COPD classification compared to each feature set alone. Methods: Participants that were never-/ever-smokers from the population-based CanCOLD study underwent spirometry and CT imaging. 107 CT PyRadiomics features were extracted from the lung, airways, vessels: 18 first-order, 14 shape, and 75 texture features. ML models including 6 demographics (age, sex, race, BMI, smoking-status, pack-years), low-attenuation-areas-below -950HU, and CT PyRadiomics were evaluated for classifying COPD (FEV1/FVC<0.70). The area under the receiver operating characteristic curve (AUC) and DeLong’s test were used to compare models. Results: 1302 participants were evaluated (n=669 no COPD; n=363 mild COPD; n=270 moderate/severe COPD). The model with CT lung parenchyma+airway+vessel features (AUC=0.87; p<0.05) obtained a higher performance than models with lung (AUC=0.78), airway (AUC=0.84), or vessel features alone (AUC=0.81). The lung+vessel (AUC=0.83) and the airway+vessel model (AUC=0.85) obtained higher performance than the lung only model (AUC=0.78; p<.05). Conclusion: Inclusion of CT features extracted from the pulmonary vessels improves classification of COPD in a mild COPD cohort.
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 | 0.002 | 0.007 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.002 | 0.001 |
| Open science | 0.000 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.006 | 0.002 |
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".