Assessment of perivascular space filtering methods using a three-dimensional computational model
Bibliographic record
Abstract
Growing interest surrounds the assessment of perivascular spaces (PVS) on magnetic resonance imaging (MRI) and their validation as a clinical biomarker of adverse brain health. Nonetheless, the limits of validity of current state-of-the-art segmentation methods are still unclear. Here, we propose an open-source three-dimensional computational framework comprising 3D digital reference objects and evaluate the performance of three PVS filtering methods under various spatiotemporal imaging considerations (including sampling, motion artefacts, and Rician noise). Specifically, we study the performance of the Frangi, Jerman and RORPO filters in enhancing PVS-like structures to facilitate segmentation. Our findings were three-fold. First, as long as voxels are isotropic, RORPO outperforms the other two filters, regardless of imaging quality. Unlike the Frangi and Jerman filters, RORPO's performance does not deteriorate as PVS volume increases. Second, the performance of all "vesselness" filters is heavily influenced by imaging quality, with sampling and motion artefacts being the most damaging for these types of analyses. Third, none of the filters can distinguish PVS from other hyperintense structures (e.g. white matter hyperintensities, stroke lesions, or lacunes) effectively, the area under precision-recall curve dropped substantially (Frangi: from 94.21 [IQR 91.60, 96.16] to 43.76 [IQR 25.19, 63.38]; Jerman: from 94.51 [IQR 91.90, 95.37] to 58.00 [IQR 35.68, 64.87]; RORPO: from 98.72 [IQR 95.37, 98.96] to 71.87 [IQR 57.21, 76.63] without and with other hyperintense structures, respectively). The use of our computational model enables comparing segmentation methods and identifying their advantages and disadvantages, thereby providing means for testing and optimising pipelines for ongoing and future studies.
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 | 0.005 | 0.014 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.003 | 0.002 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.002 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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".