Averaged subtracted polarization imaging for endoscopic diagnostics of surface microstructures on translucent mucosae

Area of Science:

• Medical Imaging
• Optical Engineering
• Biomedical Optics

Background:

• Endoscopic visualization of mucosal microstructures is crucial for disease diagnosis.
• Conventional imaging techniques struggle with translucent tissues, limiting microstructure visibility.
• Polarization-based imaging offers potential for enhanced contrast and detail.

Purpose of the Study:

• To develop and evaluate a novel endoscopic image processing technique for enhancing microstructures on translucent mucosae.
• To implement and validate polarimetric endoscopic systems for real-time microstructure enhancement.
• To compare the proposed method against conventional color intensity image processing.

Main Methods:

• Utilizing co- and cross-polarization images under linearly polarized light to calculate the averaged subtracted polarization image (AVSPI).
• Conducting experiments with phantoms and excised tissues using a manual setup with ring lighting and polarizers.
• Developing an objective evaluation method based on texture analysis for enhanced images.
• Introducing two online, rigid-type, polarimetric endoscopic implementations with different polarimetric cameras.

Main Results:

• The AVSPI technique significantly improved microstructure enhancement compared to conventional methods.
• Texture analysis confirmed the effectiveness of the proposed method in evaluating enhanced images.
• Online polarimetric endoscopes demonstrated robust microstructure enhancement irrespective of device-specific extinction ratios.
• Both beam-splitter and single-chip monochrome polarimetric camera systems proved effective.

Conclusions:

• Polarimetric endoscopy utilizing AVSPI is a highly effective technique for enhancing mucosal microstructures.
• The developed method is practical for hardware implementation in clinical settings.
• This advancement holds promise for improved endoscopic diagnostics and interventions.