Standardization of an AI-based vocal fold assessment on an example of Recurrent Respiratory Papillomatosis model| ACTA Otorhinolaryngologica Italica

Abstract

Objective

Advancements in adjuvant intralesional treatment for recurrent respiratory papillomatosis (RRP) with bevacizumab necessitate objective methods to compare pre- and post-treatment RRP extent.

Methods

This study aims to evaluate the efficacy of an artificial intelligence (AI)-based annotation system, Glottis Coverage - Artificial Intelligence and Deep learning (GC-AID), for assessing affected mucosa in white light and narrow band imaging modalities, a case-study for future applications.

Results

In healthy larynxes, the mean difference between areas of the vocal folds was 2.64%. For patient #4, following the bevacizumab injection, RRP coverage in white light decreased from 69.53% to 42.63%. A similar improvement was observed for patient #1, while no significant benefits were noted for patients #2 and #3.

Conclusion

The reduction of the extent of RRP was precisely measured with GC-AID tool. Obtaining such objectivized, quantitative results was possible with frame extraction and annotation using the NBI-ML system that was developed by us.