Evaluation der Primärbefundung durch ein CAD-System in der Mammadiagnostik
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
PURPOSE: To assess the capability of the computer assisted detection (CAD) system to classify calcifications that are histologically verified as malignant and benign or are proven benign by magnification and follow up mammography. MATERIALS AND METHODS: Three groups of microcalcifications (MC) with and without associated masses were enrolled in the study. The cancer group included 141 screen-detected breast cancer cases. One benign group comprised 109 cases with histologically benign specimens obtained through a minimally invasive breast biopsy. A second benign group included 72 lesions with MC that appeared benign on magnification/compression views and were confirmed to be benign on follow-up mammograms over a period of at least 1.5 years. All mammograms were evaluated with a CAD system (Second Look version 3.5, CADx Medical Systems, Canada). RESULTS: CAD correctly detected 125 of 141 (89 %) cancer cases. Of the 16 false negative cases, CAD marked the location of the MC (which were associated with malignant mass) with a mass mark in 12 cases. For benign cases, CAD did not correctly mark the microcalcifications in 59 of the 109 lesions confirmed benign histologically (54.1 %) and in 39 of the 72 lesions established benign mammographically (54.2 %). Adenosis introduced the highest rate of falsely marked microcalcifications (62 %). CONCLUSION: Due to its limited specificity, CAD can still not be recommended for the primary classification of microcalcifications as malignant or benign. Nevertheless, the low false negative rate and rather high detection rate of malignant findings indicate some value of CAD for an independent second reading.
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.
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 | 0.007 | 0.001 |
| Meta-epidemiology (narrow) | 0.003 | 0.003 |
| Meta-epidemiology (broad) | 0.003 | 0.001 |
| Bibliometrics | 0.002 | 0.003 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.003 | 0.003 |
| Insufficient payload (model declined to judge) | 0.003 | 0.003 |
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 it