Enhanced volumetric visualization for real time 4D intraoperative ophthalmic swept-source OCT

Area of Science:

• Medical Imaging
• Computer Graphics
• Biomedical Engineering

Background:

• Current rendering software for volumetric Optical Coherence Tomography (OCT) data presents indistinct tissue features and poor depth perception.
• Real-time, high-quality visualization is crucial for intrasurgical OCT, especially with advancements in hand-held and microscope-integrated devices.

Purpose of the Study:

• To develop and demonstrate an enhanced, real-time volumetric rendering pipeline for OCT data.
• To improve the visual appeal, speed, and depth perception of OCT visualizations for surgical guidance.

Main Methods:

• Implemented an integrated volumetric rendering pipeline using ray casting on Graphics Processing Unit (GPU) hardware.
• Incorporated high-performance volumetric median and Gaussian filtering, boundary/feature enhancement, depth encoding, and lighting.
• Tested the pipeline on tissue phantoms and during live human surgical imaging.

Main Results:

• The enhanced pipeline achieved real-time volumetric rendering of OCT data.
• Visualizations demonstrated significantly improved clarity of tissue features and depth perception compared to standard ray casting.
• The system successfully rendered 3D visualizations of pathology and intraoperative maneuvers.

Conclusions:

• The developed GPU-accelerated rendering pipeline offers a substantial improvement for real-time OCT data visualization.
• This enhanced visualization supports better surgical guidance and understanding of intraoperative anatomy and pathology.
• The pipeline's performance and visual quality meet the demands of modern intrasurgical OCT applications.