Skip to main content

Table 1 Main parameters of the three datasets used in this work

From: Generalizable attention U-Net for segmentation of fibroglandular tissue and background parenchymal enhancement in breast DCE-MRI

 

Dataset 1

Dataset 2

Dataset 3

Dataset 4

Institution

Our institution

Our institution

Our institution

Duke dataset

Acquisition time range

September 2013–October 2015

September 2020–October 2022

June 2018–October 2022

January 2000–March 2014

Magnetic field strength [T]

3.0

1.5

3.0

3.0

Manufacturer and model

Siemens, Magnetom Skyra

Siemens, Sola

Siemens, Magnetom Skyra

Siemens, Magnetom Skyra

Resolution [px × px]

448 × 448

384 × 384

448 × 448

448 × 448

TR [ms]

4.40–4.64

4.98

4.40–4.48

3.77

TE [ms]

1.56–1.78

2.39

1.62–1.70

1.44

Flip angle [°]

10

10

10

10

FGT model: patients (cases)/volumes/slices

29/29/3048

4/4/436

71/76/9088

3/3/672

BPE model: patients (cases)/volumes/slices

82/169/17420

4/4/1744

71/76/9088

3/6/1344

Patients’ age mean ± std [years]

47.3 ± 11.6

40.3 ± 19.4

48.8 ± 12.6

52.6 ± 6.6

Use in the study

Patient stratified train/valid/test split

Test set

Volumetric analysis

Volumetric analysis

Test set