Atlas-based automatic planning and 3D-2D fluoroscopic guidance in pelvic trauma surgery

Area of Science:

• Medical Imaging
• Surgical Planning
• Computational Anatomy

Background:

• Percutaneous screw fixation for pelvic trauma is complex, often involving prolonged fluoroscopy and multiple insertion attempts.
• Navigating narrow pelvic bone corridors requires high precision to avoid complications.

Purpose of the Study:

• To develop an automated method for planning surgical trajectories in pelvic trauma surgery using preoperative CT scans.
• To assist device placement by augmenting fluoroscopic images with planned trajectories.

Main Methods:

• Constructed a statistical shape model (SSM) from 40 CT images to define safe trajectory regions within pelvic bone corridors.
• Utilized active shape model (ASM) registration and free-form deformation (FFD) to map the SSM to patient-specific CT scans.
• Augmented intraoperative fluoroscopy with projected planned trajectories via 3D-2D registration.

Main Results:

• Leave-one-out cross-validation showed surface RMSE of 1.8 ± 0.2 mm after FFD refinement.
• Automatically planned trajectories demonstrated >90% conformance within bone corridors (PPV) and were 3-5 mm from the bone cortex.
• Cadaver studies confirmed accurate 3D-2D registration (<4° rotation) and >90% overlay conformance in augmented fluoroscopy.

Conclusions:

• The developed method enables automatic planning of pelvic fracture fixation and enhances fluoroscopy for improved surgical precision and safety.
• This approach does not require CT segmentation or additional hardware, integrating seamlessly into existing workflows.
• Potential benefits include reduced operating time and radiation dose by minimizing iterative screw placement attempts.