Real time 3D visualization of intraoperative organ deformations using structured dictionary.

Area of Science:

• Medical Imaging
• Computer-Aided Surgery
• Surgical Navigation

Background:

• Minimally invasive surgery (MIS) faces challenges with restricted visualization of the surgical field.
• Current systems lack high-resolution, real-time 3D visualization for anatomical recognition during MIS.
• Accurate intraoperative visualization is crucial for safe MIS procedures.

Purpose of the Study:

• To develop a real-time 3D visualization approach for organ deformations in MIS.
• To improve the recognition of anatomical structures using limited optical imaging and preoperative scans.
• To address the limitations of current visualization systems in MIS.

Main Methods:

• A novel reconstruction method utilizing optical imaging patches with a limited field-of-view.
• Integration of a single preoperative scan (MRI or CT) with optical data.
• Utilizing spherical harmonic coefficients within structured dictionaries for surface reconstruction.
• Registration of partial optical images with preoperative scans.

Main Results:

• Achieved sub-3 mm spatial resolution (Hausdorff distance) using a single MRI scan and limited optical view.
• Demonstrated real-time reconstruction frame rates between 10-39 frames/s.
• Validated the method with finite element modeling and ex vivo experiments on porcine kidneys.

Conclusions:

• The proposed method offers a viable solution for real-time 3D organ deformation visualization in MIS.
• It enhances intraoperative visualization by integrating limited optical data with preoperative scans.
• This approach has the potential to improve safety and precision in minimally invasive procedures.