A Comparison of Hole-filling Methods for 3D Medical Data Reconstruction and Visualization

Authors

  • Chan Vei Siang School of Computing Faculty of Engineering Universiti Teknologi Malaysia 81310 UTM Johor Bahru, Johor, Malaysia
  • Farhan Mohamed School of Computing Faculty of Engineering Universiti Teknologi Malaysia 81310 UTM Johor Bahru, Johor, Malaysia

DOI:

https://doi.org/10.11113/ijic.v9n2.236

Keywords:

Medical imaging, 3D reconstruction, Sscientific visualization, ultrasound, CT data set, pixel nearest neighbor, hole-filling method

Abstract

3D medical imaging can help the physicians to understand the patient anatomy, such as 3D ultrasound and 3D CT scan. In the case of 3D ultrasound reconstruction, the pixel nearest neighbour is one of the popular methods used. The hole-filling method in pixel nearest neighbour can fill in the empty voxel data which is not captured by the ultrasound. Thus, this paper studied the various hole-filling methods in the pixel nearest neighbour method to reconstruct the missing voxels. Besides that, an alternative method is also introduced based on the modified butterfly interpolation scheme. The experiment setup is designed to test the efficiency of the hole-filling method for 3D medical data visualisation by using mean absolute error as well as qualitatively compare the visualization of the reconstruction results. The proposed method can extract smooth skin from the reconstruction volume, although it has a high average MAE result.

Author Biographies

Chan Vei Siang, School of Computing Faculty of Engineering Universiti Teknologi Malaysia 81310 UTM Johor Bahru, Johor, Malaysia

School of Computing, Faculty of Engineering, Universiti Teknologi Malaysia (UTM)

Farhan Mohamed, School of Computing Faculty of Engineering Universiti Teknologi Malaysia 81310 UTM Johor Bahru, Johor, Malaysia

School of Computing, Faculty of Engineering, Universiti Teknologi Malaysia (UTM)

Downloads

Published

2019-11-28

How to Cite

Vei Siang, C., & Mohamed, F. (2019). A Comparison of Hole-filling Methods for 3D Medical Data Reconstruction and Visualization. International Journal of Innovative Computing, 9(2). https://doi.org/10.11113/ijic.v9n2.236

Issue

Section

Computer Science