Reader accuracy and confidence in diagnosing diffuse lung disease on high-resolution computed tomography of the lungs: impact of sampling frequency
Bibliographic record
Abstract
BACKGROUND: The accuracy of the number of high-resolution computed tomography (HRCT) images necessary to diagnose diffuse lung disease (DLD) is not well established. PURPOSE: To evaluate the impact of HRCT sampling frequency on reader confidence and accuracy for diagnosing DLD. MATERIAL AND METHODS: HRCT images of 100 consecutive patients with proven DLD were reviewed. They were: 48 usual interstitial pneumonia, 22 sarcoidosis, six hypersensitivity pneumonitis, five each of desquamative interstitial pneumonitis, eosinophilic granulomatosis, and lymphangioleiomyomatosis, and nine others. Inspiratory images at 1-cm increments throughout the lungs and three specified levels formed complete and limited examinations. In random order, three experts (readers 1, 2, and 3) ranked their top three diagnoses and rated confidence for their top diagnosis, independently and blinded to clinical information. RESULTS: Using the complete versus limited examinations for correct first-choice diagnosis, accuracy for reader 1 (R1) was 81% versus 80%, respectively, for reader 2 (R2) 70% versus 70%, and for reader 3 (R3) 64% versus 59%. Reader accuracy within their top three choices for complete versus limited examinations was: R1 91% versus 91% of cases, respectively, R2 84% versus 83%, and R3 79% versus 72% of cases. No statistically significant differences were found between the diagnosis methods (P=0.28 for first diagnosis and P=0.17 for top three choices). The confidence intervals for individual raters showed considerable overlap, and the point estimates are almost identical. The mean interreader agreement for complete versus limited HRCT for both top and top three diagnoses were the same (moderate and fair, respectively). The mean intrareader agreement between complete and limited HRCT for top and top three diagnoses were substantial and moderate, respectively. CONCLUSION: Overall reader accuracy and confidence in diagnosis did not significantly differ when fewer or more HRCT images were used.
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
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 teacher head, 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".