Artificial intelligence and CT-based 3D statistical modeling to assess transsacral corridors and plan implant...

Area of Science:

• Orthopedic Surgery
• Medical Imaging
• Artificial Intelligence

Background:

• Transsacral corridors at S1 and S2 are crucial for percutaneous fixation of sacral fragility fractures.
• Ensuring implants are completely intraosseous is vital for safety, but standard imaging lacks detailed corridor visualization.
• Accurate 3D assessment of these osseous spaces is needed for precise surgical planning.

Purpose of the Study:

• To develop an AI-driven workflow for the 3D assessment of transsacral corridors (S1 and S2).
• To facilitate the precise planning of transsacral implant positioning for fragility fractures.
• To improve the safety and accuracy of percutaneous fixation techniques.

Main Methods:

• Computed a 3D statistical model of the pelvic ring from 100 pelvic CT scans.
• Trained regression learners using morphologic features to predict personalized 3D pelvic models.
• Matched models to 20 CTs of patients with sacral fragility fractures for corridor analysis.

Main Results:

• The AI models successfully demonstrated the availability, dimensions, cross-section, and symmetry of S1 and S2 transsacral corridors.
• Planned implant positions, dimensions, axes, and entry/exit points were accurately determined.
• The complete intraosseous pathway for implant placement was confirmed via CT reconstructions.

Conclusions:

• An AI and 3D statistical modeling workflow was established for determining transsacral corridors S1 and S2.
• This method enables accurate 3D assessment and planning for transsacral implant fixation.
• The developed workflow enhances the precision and safety of treating sacral fragility fractures.