Table 2 Domain non-specific (DNS) pre-processing pipeline, which does not preserve anatomical structure information.

From: The impact of pre-processing techniques on deep learning breast image segmentation

Order

Parameter

2D mammography models

3D breast MRI models

1st

Orientation

No standardization

No standardization

2nd

Pixel Spacing

No standardization

No standardization

3rd

Pixel intensity

Minimum–maximum, percentiles,

Z-Score, Histogram Normalization

Minimum-Maximum, Percentiles,

Z-Score, Histogram Normalization

4th

Image Size

256 × 256, 512 × 512, 1024 × 1024

256 × 448, 512 × 896, 1024 × 1792

256 × 256 × 64, 512 × 512 × 64

  1. The table shows the order of pre-processing steps and the corresponding methods evaluated for two datasets composed of: (a) 2D mammography and (b) 3D breast MRI. The steps include image resizing, pixel intensity normalization (Min-Max, P1-P99, Z-Score, HistNorm), pixel spacing, and orientation.