Abstract 3208: Sensitivity and Reliability of SWI Compared to T2* GRE MRI for Detection of Microbleeds in Cerebral Amyloid Angiopathy
Bibliographic record
Abstract
Introduction: Case examples and small case series suggest that MRI susceptibility weighted imaging (SWI) may be more sensitive for cerebral microbleed (CMB) detection compared to MRI T2* gradient-recalled echo (GRE). However, there are few data on CMB counts measured by SWI vs. GRE, or inter-rater reliability, in groups of patients with cerebral small vessel disease. We used data from a prospective cohort study of cerebral amyloid angiopathy (CAA), a cerebral small-vessel disease marked by high numbers of CMBs, to quantify the sensitivity and reliability of SWI vs. GRE for CMB detection. Methods: Nine patients with symptomatic CAA (mean age 71±8.3; 7 males and 2 females) and 21 healthy non-CAA controls (mean age 68±6.3; 10 M/11 F) underwent T2* GRE and SWI on a 3.0T MR scanner. Probable CAA was diagnosed according to the Boston criteria prior to study entry using information from clinical MRI with GRE sequences. Two raters (labeled 1 and 2) independently interpreted the GRE and SWI scans blinded to clinical information. The phase-filtered magnitude image was used for SWI interpretation. Agreement reliability was assessed using the kappa coefficient (where a kappa of ≥0.60 indicates good agreement) or the intraclass correlation coefficient (ICC). Results: Overall, the raters identified 1,432 CMBs in the 9 CAA cases (range 1-434 per patient) and 8 CMBs in the healthy controls (range 0-3). Rater 1 identified CMBs in 5/21 healthy controls on SWI and 5/21 on GRE, while rater 2 identified CMBs in 4/21 on SWI and 3/21 on GRE (kappa 0.70 for GRE and 0.57 for SWI). In CAA cases more CMBs were seen on SWI compared to the GRE sequence but the difference was significant only for rater 1 (rater 1: on average 85% more per patient on SWI than on GRE, p=0.008; rater 2: 19% more, p=0.25). Among CAA cases the reliability between raters was poor for GRE (ICC 0.36) but excellent for SWI (0.94, p<0.05 for comparison with GRE). Review suggested that the differing reliability was because rater 1 was less likely than rater 2 to identify faint lesions on GRE as CMB, whereas these lesions were more conspicuous on SWI. If SWI rather than GRE were used to determine CAA status according to the Boston criteria, all 9 CAA cases would remain classified as probable CAA but 2/21 controls would be reclassified as either possible (n=1) or probable (n=1) asymptomatic CAA based on the detection of one or more lobar microbleeds on SWI. Conclusions: SWI confers greater reliability as well as greater sensitivity for CMB detection compared to GRE, and should be the preferred sequence for quantifying CMBs. SWI may more frequently identify lobar microbleeds that could represent asymptomatic CAA. Further research is needed to determine whether the Boston criteria require revision to take into account the greater sensitivity of SWI for CMB detection.
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 | 0.007 | 0.019 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.000 | 0.001 |
| Research integrity | 0.001 | 0.000 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
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".