Skip to main content
Visual Computing for Industry, Biomedicine, and Art

Table 1 Comparison of DVF errors between SMEIR, SMEIR-Bio and SMEIR-Unet techniques

From: Advanced 4-dimensional cone-beam computed tomography reconstruction by combining motion estimation, motion-compensated reconstruction, biomechanical modeling and deep learning

(Mean ± SD) DVF Error (mm)

SMEIR

SMEIR-Bio

SMEIR-Unet

X

0.79 ± 0.89

0.57 ± 0.78

1.00 ± 0.98

Y

1.55 ± 0.77

0.86 ± 0.62

0.61 ± 0.82

Z

4.31 ± 1.07

2.53 ± 1.00

2.77 ± 0.99

Vector

4.98 ± 1.11

3.11 ± 1.03

3.30 ± 1.08

  1. X, Y, and Z indicate the DVF errors along the three Cartesian directions in space, respectively. The fifth row shows the vector length of the DVF error. SMEIR Simultaneous motion estimation and image reconstruction; DVF Deformation-vector-field; SMEIR-Bio Biomechanical modeling-guided SMEIR; SMEIR-Unet SMEIR with U-net based DVF fine-tuning.
Back to article page