Shape-Based Acetabular Cartilage Segmentation: Application to CT and MRI Datasets

Shape-based acetabular cartilage segmentation: application to CT and MRI datasets

Pooneh R. Tabrizi1*, Reza A. Zoroofi1, Futoshi Yokota2, Takashi Nishii3, and Yoshinobu Sato2

1 Control and Intelligent Processing Center of Excellence, School of Electrical and Computer Engineering, University of Tehran, Tehran, IRAN.

2 Imaging-based Computational Biomedicine (ICB) Lab, Graduate School of Information Science of Nara Institute of Science and Technology (NAIST),

Osaka, 565–015141, Japan.

3 Graduate School of Medicine, Osaka University Yamadaoka 2–2, Suita-shi, Osaka, 565–015141, Japan.

E-mail:

Our improved method
The method proposed in [7]
The method proposed in [22]
The method proposed in [14]

Online Resource1 Accuracy levels (1st column) and box plots (2nd column) for pelvic bone segmentation results in CT arthrography datasets with the higher radiation dose using DICE, and ASSD (mm) criteria: Black and white lines in each box indicate the median and mean values, respectively

Our improved method
The method proposed in [7]
The method proposed in [22]
The method proposed in [14]

Online Resource2 Accuracy levels (1st column) and box plots (2nd column) for pelvic bone segmentation results in CT arthrography datasets with the lower radiation dose using DICE, and ASSD (mm) criteria: Black and white lines in each box indicate the median and mean values, respectively

Our improved method
The method proposed in [7]
The method proposed in [22]
The method proposed in [15]

Online Resource3 Accuracy levels (1st column) and box plots (2nd column) for pelvic bone segmentation results in MRI datasets using DICE, and ASSD (mm) criteria: Black and white lines in each box indicate the median and mean values, respectively

Combination of the K-OPLS and Graph-Cut methods based on the manual pelvic bone segmentation method
Combination of the K-OPLS and Graph-Cut methods based on our improved pelvic bone segmentation method
The method proposed in [7]based on the manual pelvic bone segmentation method
The method proposed in [7] based on our improved pelvic bone segmentation method
The method proposed in [14]
The method proposed in [13]

Online Resource4 Accuracy levels (1st column) and box plots (2nd column) for cartilage segmentation results in CT arthrography datasets with the higher radiation dose using DICE, CTMEmean (mm), and CTMEmax (mm) criteria: Black and white lines in each box indicate the median and mean values, respectively

Combination of the K-OPLS and Graph-Cut methods based on the manual pelvic bone segmentation method
Combination of the K-OPLS and Graph-Cut methods based on our improved pelvic bone segmentation method
The method proposed in [7] based on the manual pelvic bone segmentation method
The method proposed in [7] based on our improved pelvic bone segmentation method
The method proposed in [14]
The method proposed in [13]

Online Resource5 Accuracy levels (1st column) and box plots (2nd column) for cartilage segmentation results in CT arthrography datasets with the lower radiation dose using DICE, CTMEmean (mm), and CTMEmax (mm) criteria: Black and white lines in each box indicate the median and mean values, respectively

Combination of the K-OPLS and Graph-Cut methods based on the manual pelvic bone segmentation method
Combination of the K-OPLS and Graph-Cut methods based on our improved pelvic bone segmentation method
The method proposed in [7] based on the manual pelvic bone segmentation method
The method proposed in [7] based on our improved pelvic bone segmentation method
The method proposed in [15]
The method proposed in [13]

Online Resource6 Accuracy levels (1st column) and box plots (2nd column) for cartilage segmentation results in MRI datasets using DICE, CTMEmean (mm), and CTMEmax (mm) criteria: Black and white lines in each box indicate the median and mean values, respectively

1