Table 2.
Summary of top 10 ranked histologic features identified in the random forest model for the prediction of global and localized % fibroglandular volume (FGV)
| Feature Name | Rank of feature importance Predicted model: global FGV (%) | Rank of feature importance Predicted model: localized FGV (%) |
|---|---|---|
| Global tissue amount | ||
| Fat amount (µm2) | 5 | 3 |
| Fat amount normalized (%) | 2 | 2 |
| Stroma amount (µm2) | 3 | 5 |
| Stroma amount normalized (%) | 1 | 1 |
| Epithelium amount normalized (%) | − | 8 |
| Morphology | ||
| Ecc epi regions (median) | 7 | − |
| Ecc epi regions (IQ) | 6 | 9 |
| Spatial arrangement of the epithelial regions (Area-Voronoi diagram) | ||
| Voronoi area (mean µm2) | 10 | − |
| Voronoi area (median µm2) | − | − |
| Voronoi area (IQ µm2) | 8 | 6 |
| Ratio epi to Voronoi (mean) | − | 7 |
| Ratio epi to Voronoi (median) | − | 4 |
| Ratio epi to non-epi (mean) | 9 | − |
| Spatial arrangement of the epithelial regions (Delaunay Triangulation) | ||
| Neighbors (SD) | 4 | 10 |
Ecc eccentricity, epi epithelial, IQ interquartile, FGV fibroglandular volume, SD standard deviation
Only features ranked within the top 10 for prediction of each, FGV density measure, are included in the table
Features are ranked numerically and sequentially from 1 to 10, with 1 representing the most important feature and 10 representing the 10th most important feature