Publication Type

Journal Article

Version

publishedVersion

Publication Date

8-2025

Abstract

Background: Early-stage diagnosis of laryngeal cancer significantly improves patient survival and quality of life. However, the scarcity of specialists in low-resource settings hinders the timely review of flexible nasopharyngoscopy (FNS) videos, which are essential for accurate triage of at-risk patients.Objective: We introduce a preliminary AI-based screening framework to address this challenge for the triaging of at-risk patients in low-resource settings. This formative research addresses multiple challenges common in high-dimensional FNS videos: (1) selecting clear, informative images; (2) deriving regions within frames that show an anatomical landmark of interest; and (3) classifying patients into referral grades based on the FNS video frames.Methods: The system includes an image quality model (IQM) to identify high-quality endoscopic images, which are then fed into a disease classification model (DCM) trained on efficient convolutional neural network (CNN) modules. To validate our approach, we curated a real-world dataset comprising 132 patients from an academic tertiary care center in the United States.Results: Based on this dataset, we demonstrated that the IQM quality frame selection achieved an area under the receiver operating characteristic curve (AUROC) of 0.895 and an area under the precision-recall curve (AUPRC) of 0.878. When using all the image frames selected by the IQM, the DCM improved its performance by 38% considering the AUROC (from 0.60 to 0.83) and 8% considering the AUPRC (from 0.84 to 0.91). Through an ablation study, it was demonstrated that a minimum of 50 good-quality image frames was required to achieve the improvements. Additionally, an efficient CNN model can achieve 2.5-times-faster inference time than ResNet50.Conclusions: This study demonstrated the feasibility of an AI-based screening framework designed for low-resource settings, showing its capability to triage patients for higher-level care efficiently. This approach promises substantial benefits for health care accessibility and patient outcomes in regions with limited specialist care in outpatient settings. This research provides necessary evidence to continue the development of a fully validated screening system for low-resource settings.

Keywords

artificial intelligence, cancer triage, deep learning, efficient neural nets, flexible nasopharyngoscopy, head and neck cancers, machine learning

Discipline

Artificial Intelligence and Robotics

Research Areas

Intelligent Systems and Optimization

Areas of Excellence

Sustainability

Publication

JMIR Formative Research

Volume

9

First Page

1

Last Page

9

ISSN

2561-326X

Identifier

10.2196/66110

Publisher

JMIR Publications

Additional URL

https://doi.org/10.2196/66110

Download

Share

COinS