INTRODUCTION
Endoscopic examination of the larynx typically focuses on the surface of
the mucosa. However, finer details of the mucosal and submucosal
vascular structures are possible due to the recent integration of image
enhancement systems using optical tools and digital technology into
conventional endoscopic systems. Inspection of mucosal pathologies by
these endoscopic enhancement systems has revealed vascular
characteristics of the inflammatory and the tumorigenic processes and
especially the neo-angiogenesis related to carcinogenesis.
One of these enhancement systems is narrow-band imaging (NBI), which
uses optically filtered light to produce high contrast images of the
vasculature for endoscopic diagnosis of malignant and precancerous
lesions (1). A corresponding study by the European Laryngological
Society (ELS), using NBI, defined the vascularization patterns
associated with laryngeal pathologies in a descriptive guideline (2).
Recently, a novel image enhancement system, Image1 S, has been
successfully used in different branches of medicine with the same aim;
to make endoscopic images clearer and vascular network identifiable (3,
4). This system is based on software that uses the spectral light
separations to produce images with different contrast characteristics.
The spectral separation is obtained from the camera console and is
enhanced by color processing algorithms.
In the present study, we attempted to evaluate the practical usefulness
of the Image1 S system for the detecting the vascular patterns
associated with benign, premalignant, and malignant lesions. Here, we
included a group of otolaryngologists with different levels of
proficiency as observers to discriminate between the vascular patterns
defined by the ELS and determined the level of accuracy and consistency
among the observers.