Comment on: Reproducibility of the scleroderma pattern assessed by wide-field capillaroscopy in subjects suffering from Raynaud's phenomenon: reply
Bibliographic record
Abstract
Sir, in our recent article [1] we tried to study the reproducibility and accuracy of scleroderma pattern assessed by capillaroscopy since very few data are available on the topic, although capillaroscopy was included in the new ACR/EULAR definition of SSc. Sabour [2] raised interesting methodological issues about our results. We agree that kappa is not the only result to take into consideration and this is why we also gave the number of concordant observations. The two points are the dependence of kappa upon the prevalence of the categories and the difference between accuracy and reproducibility. When we study a method such as capillaroscopy, limitations exist from a practical point of view. First, the ideal prevalence of categories should be the same as in current practice. This prevalence is highly variable depending on recruitment. The prevalence of SSc in patients consulting for RP is ∼10% in our country, but most patients are not referred for capillaroscopy and we do not know the percentage of scleroderma pattern in patients who have this examination. Thus we tried to study a significant number of positive and negative capillaroscopies selected from real life. We enrolled a large number of observers in order to reflect clinical practice. Second, precision was studied in comparison with a reference defined as a majority of concordant observations, because we have no gold standard for the interpretation of capillaroscopy. Sabour raises true methodological limitations that are unfortunately unavoidable on this subject.
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 | 0.004 | 0.033 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.002 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.002 | 0.002 |
| Scholarly communication | 0.002 | 0.003 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.042 | 0.028 |
| Insufficient payload (model declined to judge) | 0.004 | 0.005 |
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".