Automatic Assessment of Lower-Limb Alignment from Computed Tomography

Area of Science:

• Orthopaedic surgery
• Medical imaging
• Biomechanical analysis

Background:

• Preoperative planning for lower-limb realignment surgery requires precise alignment measurements.
• Traditional 2D radiographs limit the accuracy and efficiency of surgical planning.
• 3D bone models from CT scans are needed for advanced orthopaedic surgery planning.

Purpose of the Study:

• To develop and validate an automated method for quantifying lower-limb alignment parameters from CT scans.
• To enable fast and accurate 3D planning of orthopaedic surgery.
• To assess the reliability and efficiency of the automated method compared to manual measurements.

Main Methods:

• CT scans of 50 lower extremities were segmented manually and automatically.
• 32 manual landmarks were placed twice to assess intraobserver reliability.
• A shape-fitting algorithm was used for automatic landmark placement.
• 25 lower-limb alignment angles were calculated from the landmarks.

Main Results:

• The automated method showed comparable differences to manual measurements (2.17 ± 1.37 mm vs. 2.01 ± 1.64 mm).
• Automated calculation took 12 minutes, significantly less than the manual method's 60 minutes.
• Good to excellent reliability was found between manual and automatic assessments for 23 of 25 angles.

Conclusions:

• An automated method for calculating lower-limb alignment from CT scans was successfully developed.
• The method is efficient, reliable, and comparable to manual assessments.
• This automated approach offers a viable alternative for preoperative surgical planning.