Importance of <scp> R <sub>2</sub> </scp> accuracy in susceptibility source separation
Bibliographic record
Abstract
Abstract Purpose To examine the importance of R 2 accuracy on independent paramagnetic and diamagnetic outputs from susceptibility source separation in the brain from two publicly available methods. Methods The effects of R 2 errors, which translate into errors, on output maps from χ‐separation and χ‐sepnet were examined using data from 11 healthy volunteers. Baseline R 2 values were determined by Bloch modeling a dual‐echo turbo spin echo decay with measured flip angles. R 2 errors were introduced from either simple exponential fitting, R 2 multiplication factors, or R 2 approximation using only . Altered R 2 maps were then used as input for the susceptibility source separation models using either default or calculated relaxometric constant. Difference maps and mean percentage errors within regions of interest (ROIs) were measured. Results Errors in R 2 , and hence , directly affected paramagnetic and diamagnetic components. χ‐sepnet was less sensitive to R 2 errors than χ‐separation and had reduced variance among subjects. χ‐sepnet susceptibility component errors did not reach more than ±20% in most ROIs for all alteration approaches. In contrast, χ‐separation, with default relaxometric constant, reached 56% susceptibility component error with −25% R 2 error input. Exponential fitting R 2 error exceeded −25%, thus, even larger component errors occurred. ‐based approximation had −25% R 2 mean error across ROIs (−18% across whole brain), yielding 57% mean susceptibility component error across ROIs. Conclusion Paramagnetic and diamagnetic outputs of susceptibility source separation methods have variable responses to R 2 error, that may occur with simple R 2 fitting or R 2 approximation, and can be strongly biased by it.
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 | 0.010 | 0.042 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.002 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| 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".