New/Enlarging T2 Lesions in a Progressive Multiple Sclerosis Trial Cohort: Computer vs Human Detection
Bibliographic record
Abstract
Background: Although detection of new/enlarging multiple sclerosis (MS) lesions is a key metric in clinical practice and research, little is known about how human and automated detection methods compare. We compared findings made by a neuroradiologist to simulate routine practice to those made with a computer-aided technique as used in a clinical trial. Methods: MRIs from a 96-week, progressive MS clinical trial with 255 participants were evaluated for new/enlarging T2 lesions by both a neuroradiologist and a semiautomated method. Readings from 887 paired scans were compared using a paired t test and inter-rater reliability κ. Selected discordant reads were subsequently reviewed by a second neuroradiologist. Results: The semiautomated method identified new/enlarging lesions on 19.7% of the paired scans, while the neuroradiologist identified lesions on 5.7%. Of the 185 paired scans with new/enlarging T2 lesions by either method, the semiautomated method detected a mean of 3.4 new lesions, while the neuroradiologist detected 0.4. Overall κ was 0.18 (poor agreement); the κ of only scans with new/enlarging lesions by either method was –0.06. When scans were categorized as active or inactive, κ was 0.18. Unblinded neuroradiologist overread of selected discordant scans found more lesions when using advanced radiology tools such as coregistration. Conclusions: A large discordance was found between new/enlarging MS lesions identified by a neuroradiologist and a semiautomated identification method. These findings may explain clinical trials reporting new lesions in patients on highly effective therapies, while they are less common in clinical practice. These findings have implications for treatment decision algorithms using MRI. Advanced radiology tools such as coregistration may improve lesion detection in routine clinical practice.
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 | 0.001 | 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".