Abstract P5-14-15: Discovery and validation of simple heart measurements to screen patients who will benefit from deep inspiration breath hold for left breast irradiation
Bibliographic record
Abstract
Abstract Purpose: Deep inspiration breath hold (DIBH) using active breath control in left-sided breast irradiation (RT) can significantly reduce radiation dose to heart and coronary arteries in selected patients. At our institution, a cutoff of V50% >10cc is currently used to identify patients who require DIBH. This dose-volume cutoff requires generation of a radiation plan in order to select patients. The purpose of this study was to identify simple 2-D measurements of the heart at the time of CT simulation that can help screen patients for DIBH in order to streamline the process for left-sided breast RT. Methods: This study evaluated CT simulation scans from 50 left-sided breast cancer patients treated with tangent RT alone, where 50% of patients were treated with DIBH. On each CT dataset, a tangential line was drawn between the medial and lateral tattoos. The following heart measurements were recorded by a blinded observer at 2, 3, 4, and 5 cm below the tattoos: (1) maximal heart distance (MHD) perpendicular to this line and (2) heart length (HL) along this line. Predictive models were created using two strategies; (1) step wise approach utilizing the most significant factor and (2) principle component analysis. Using C-statistics, 3 cutoffs with the highest correlations with heart dose were determined from the model. These cutoffs were validated in a cohort of 100 consecutive patients treated from Jan-May 2012. Results: Data from 49 patients in the derivation cohort were analyzable. The HL at 2 cm had the strongest correlation with V50 (R2 = 0.45; p <0.0001). Other values that were associated with V50 were HL at 3 cm (R2 = 0.37; p = <0.0001), MHD at 2cm (R2 = 0.25; p = 0.0003), MHD at 3 cm (R2 = 0.23; p = 0.0006) and HL at 4 cm (R2 = 0.17, p = 0.0035). The predictive model using HL at 2 cm gave an adjusted R2 = 0.4385 (P<0.0001). Adding other variables into the predictive model did not improve the adjusted R2. The following 3 cutoffs for HL at 2 cm with respective (sensitivity; specificity) were identified: 3.92 cm (0.955; 0.48), 7.65 cm (0.32; 0.48), and 6.5 cm (0.5; 0.8). In the validation cohort, 22/100 had V50% > 10 cc. Validation of the 3 derived cutoffs for HL at 2 cm yielded the following sensitivities and specificities: 1.0 and 0.28 for 3.92 cm; 0.53 and 0.93 for 7.65 cm; 0.58 and 0.69 for 6.5 cm. Positive predictive value (PPV) and negative predictive value (NPV) were as follows: 3.92 cm (PPV = 0.25; NPV = 1.00). 7.65 cm (PPV = 0.63; NPV = 0.89), and 6.5cm (PPV = 0.30; NPV = 0.87). Conclusions: A simple 2-D heart measurement, heart length at 2 cm below the tattoos, had moderate correlation with the irradiated heart volume. Although HL at 2 cm did not have both high specificity and sensitivity to predict who would benefit from DIBH, cut-offs could be selected to maximize sensitivity (3.92 cm for sensitivity 1.0) or specificity (7.65 cm for specificity 0.93). For example, in departments that routinely acquire both free-breathing and DIBH CT simulation scans for all left-sided patients, a cutoff of 3.92cm would minimize the number of unnecessary DIBH scans. Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P5-14-15.
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 | 0.003 | 0.010 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| 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".