Table 3 Quantitative evaluation of DBAHNet segmentation performance for the secondary datasets. The results include Dice Similarity Coefficient (DSC) and 95th percentile of Hausdorff Distance (HD95) for both cortical (C) and trabecular (T) compartments. \(N_{\text {sub}}\) represents the total number of subsets, with each scan divided into 10 subsets along the z-axis.

From: A robust deep learning approach for segmenting cortical and trabecular bone from 3D high resolution µCT scans of mouse bone

Datasets

Avg DSC

DSC C

DSC T

Avg HD95

HD95 C

HD95 T

Dataset 437 (\(N_{\text {sub}}\) = 40)

0.9620

0.9963

0.9270

0.0342

0.0071

0.0612

Dataset 538 (\(N_{\text {sub}}\) = 26)

0.7691

0.8773

0.6609

0.1365

0.0935

0.1796

Dataset 639 (\(N_{\text {sub}}\) = 37)

0.9494

0.9754

0.9233

0.0155

0.0104

0.0156

Dataset 740 (\(N_{\text {sub}}\) = 40)

0.9110

0.9731

0.8489

0.0215

0.0154

0.0278

All datasets

Mean performance

0.8979

0.9555

0.8395

0.0520

0.0311

0.0710

Back to article page