Multicenter validation of automated detection of paramagnetic rim lesions on brain MRI in multiple sclerosis
Bibliographic record
Abstract
BACKGROUND AND PURPOSE: Paramagnetic rim lesions (PRLs) are an MRI biomarker of chronic inflammation in people with multiple sclerosis (MS). PRLs may aid in the diagnosis and prognosis of MS. However, manual identification of PRLs is time-consuming and prone to poor interrater reliability. To address these challenges, the Automated Paramagnetic Rim Lesion (APRL) algorithm was developed to automate PRL detection. The primary objective of this study is to evaluate the accuracy of APRL for detecting PRLs in a multicenter setting. METHODS: We applied APRL to a multicenter dataset, which included 3-Tesla MRI acquired in 92 participants (43 with MS, 14 with clinically isolated syndrome [CIS]/radiologically isolated syndrome [RIS], 35 without RIS/CIS/MS). Subsequently, we assessed APRL's performance by comparing its results with manual PRL assessments carried out by a team of trained raters. RESULTS: Among the 92 participants, expert raters identified 5637 white matter lesions and 148 PRLs. The automated segmentation method successfully captured 115 (78%) of the manually identified PRLs. Within these 115 identified lesions, APRL differentiated between manually identified PRLs and non-PRLs with an area under the curve (AUC) of .73 (95% confidence interval [CI]: [.68, .78]). At the subject level, the count of APRL-identified PRLs predicted MS diagnosis with an AUC of .69 (95% CI: [.57, .81]). CONCLUSION: Our study demonstrated APRL's capability to differentiate between PRLs and lesions without paramagnetic rims in a multicenter study. Automated identification of PRLs offers greater efficiency over manual identification and could facilitate large-scale assessments of PRLs in clinical trials.
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 | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| 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".