Fully automated side branch detection in intravascular optical coherence tomography pullback runs

Area of Science:

• Medical Imaging
• Cardiovascular Research
• Image Analysis

Background:

• Side branches in atherosclerotic lesions significantly impact treatment strategies and serve as landmarks for image registration.
• Intravascular optical coherence tomography (IVOCT) offers high-resolution coronary morphology, making it valuable for side branch analysis.

Purpose of the Study:

• To develop and validate a fully automated method for detecting side branches in IVOCT images.
• To precisely segment key structures including the imaging catheter, protective sheath, guide wire, and lumen.

Main Methods:

• A fully automated algorithm was developed for side branch detection in IVOCT images.
• The method involved precise segmentation of the imaging catheter, protective sheath, guide wire, and lumen.
• Validation was performed using 25 in-vivo datasets.

Main Results:

• High intraclass correlation coefficients (0.997, 0.949, 0.974) were achieved for catheter, sheath, and lumen segmentation.
• Guide wires were detected with 0.97 Dice's coefficient for shadow regions.
• The algorithm detected 94.0% of side branches with 5.0% false positives and a 0.85 Dice's coefficient for side branch size.

Conclusions:

• The presented automated method is accurate and robust for side branch analysis in IVOCT images.
• This technique can aid in optimizing treatment strategies and improving image registration in cardiovascular interventions.