Three-Dimensional Segmentation of Equine Paranasal Sinuses in Multidetector Computed Tomography Datasets: Preliminary...

Area of Science:

• Veterinary Anatomy
• Equine Imaging
• Biomedical Engineering

Background:

• The equine paranasal sinus system is complex and bilaterally symmetrical.
• Understanding age-related anatomical variations is crucial for equine health and research.

Purpose of the Study:

• To extract morphometric measures from CT images of equine heads.
• To implement clustering analysis for computer-aided identification of age-related variations in paranasal sinuses.

Main Methods:

• CT imaging of 18 equine cadaver heads (2-25 years old).
• Segmentation to extract volume, surface area, and relative density of eight paranasal sinuses.
• K-means clustering analysis on age-grouped data (young, middle-aged, old).

Main Results:

• Morphometric measurements varied by sinus position and horse age, not body side.
• Volume and surface area of the ventral conchal sinus (VCS), rostral maxillary sinus (RMS), and caudal maxillary sinus (CMS) increased with age.
• Age-related clustering was confirmed for RMS (0.72 accuracy) and CMS (0.67 accuracy), but not for VCS (0.31 accuracy).

Conclusions:

• CT-based 3D imaging and clustering can identify age-related variations in specific equine paranasal sinuses (RMS, CMS).
• This methodology provides a foundation for computer-aided diagnosis and research in equine sinus anatomy.