Anatomical noise in contrast‐enhanced digital mammography. Part I. Single‐energy imaging
Bibliographic record
Abstract
PURPOSE: Dual-energy (DE) contrast-enhanced digital mammography (CEDM) uses an iodinated contrast agent in combination with digital mammography (DM) to evaluate lesions on the basis of tumor angiogenesis. In DE imaging, low-energy (LE) and high-energy (HE) images are acquired after contrast administration and their logarithms are subtracted to cancel the appearance of normal breast tissue. Often there is incomplete signal cancellation in the subtracted images, creating a background "clutter" that can impair lesion detection. This is the second component of a two-part report on anatomical noise in CEDM. In Part I the authors characterized the anatomical noise for single-energy (SE) temporal subtraction CEDM by a power law, with model parameters α and β. In this work the authors quantify the anatomical noise in DE CEDM clinical images and compare this with the noise in SE CEDM. The influence on the anatomical noise of the presence of iodine in the breast, the timing of imaging postcontrast administration, and the x-ray energy used for acquisition are each evaluated. METHODS: The power law parameters, α and β, were measured from unprocessed LE and HE images and from DE subtracted images to quantify the anatomical noise. A total of 98 DE CEDM cases acquired in a previous clinical pilot study were assessed. Conventional DM images from 75 of the women were evaluated for comparison with DE CEDM. The influence of the imaging technique on anatomical noise was determined from an analysis of differences between the power law parameters as measured in DM, LE, HE, and DE subtracted images for each subject. RESULTS: In DE CEDM, weighted image subtraction lowers β to about 1.1 from 3.2 and 3.1 in LE and HE unprocessed images, respectively. The presence of iodine has a small but significant effect in LE images, reducing β by about 0.07 compared to DM, with α unchanged. Increasing the x-ray energy, from that typical in DM to a HE beam, significantly decreases α by about 2×10(-5) mm2, and lowers β by about 0.14 compared to LE images. A comparison of SE and DE CEDM at 4 min postcontrast shows equivalent power law parameters in unprocessed images, and lower α and β by about 3×10(-5) mm2 and 0.50, respectively, in DE versus SE subtracted images. CONCLUSIONS: Image subtraction in both SE and DE CEDM reduces β by over a factor of 2, while maintaining α below that in DM. Given the equivalent α between SE and DE unprocessed CEDM images, and the smaller anatomical noise in the DE subtracted images, the DE approach may have an advantage over SE CEDM. It will be necessary to test this potential advantage in future lesion detectability experiments, which account for realistic lesion signals. The authors' results suggest that LE images could be used in place of DM images in CEDM exam interpretation.
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 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| 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".