Evaluation of Software Tools for Segmentation of Temporal Bone Anatomy

Area of Science:

• Medical imaging
• Surgical simulation
• Anatomical modeling

Background:

• 3D medical image data is crucial for surgical planning.
• Virtual 3D models of complex anatomy, like the temporal bone, aid surgical education and rehearsal.
• Accurate segmentation of anatomical structures from medical images is essential for creating these models.

Purpose of the Study:

• To evaluate the efficacy of publicly available software packages for segmenting temporal bone anatomy.
• To determine the best approach for creating patient-specific virtual temporal bone models for surgical simulation.

Main Methods:

• Four software packages (ITK-SNAP, MITK, 3D Slicer, Seg3D) were assessed for temporal bone segmentation.
• Segmentation was performed on CT and MR images.
• The usability and efficiency of each tool were evaluated.

Main Results:

• No single software package could efficiently segment all temporal bone structures.
• A combination of the evaluated tools facilitated the creation of a complete virtual temporal bone model.
• The process required reasonably minimal effort when using multiple applications.

Conclusions:

• A multi-software approach is recommended for comprehensive temporal bone segmentation.
• This strategy supports the creation of detailed virtual models for patient-specific surgical simulation.
• Combining ITK-SNAP, MITK, 3D Slicer, and Seg3D offers a viable workflow for anatomical modeling.