Confidence-guided integration of otoscopy videos and tympanometry for middle ear disease diagnosis: A multi-center otoscopy study with external paired-cohort evaluation
5 affiliations
- 1Center for Artificial Intelligence Research, Wake Forest University School of Medicine, Winston-Salem, North Carolina, United States of America
- 2Department of Biostatistics and Data Science, Wake Forest University School of Medicine, Winston-Salem, North Carolina, United States of America
- 3Department of Otolaryngology—Head and Neck Surgery, Nationwide Children’s Hospital and The Ohio State University, Columbus, Ohio, United States of America
- 4Department of Otolaryngology-Head and Neck Surgery, Vanderbilt University Medical Center, Nashville, Tennessee, United States of America
- 5Department of Pathology, The Ohio State University, Columbus, Ohio, United States of America
Abstract
Accurate diagnosis of middle ear diseases using otoscopy remains challenging, particularly in primary care settings where clinician experience with otoscopy varies and visual examination alone may not fully capture middle ear physiology. Although AI-based models have shown promise for automated otoscopy interpretation, most rely primarily on visual information, creating an opportunity to improve diagnostic robustness by integrating complementary physiological measurements. Here, we present OtoTymp-AI, a confidence-guided multimodal decision-fusion framework that integrates otoscopy video analysis with conventional 226 Hz tympanometry. A convolutional neural network was trained to classify otoscopy videos, and tympanometric measurements were incorporated through a clinically interpretable, rule-guided decision strategy to refine visually uncertain predictions rather than through jointly trained multimodal representation learning. Within this multi-center otoscopy study, multimodal evaluation was performed in one independent external cohort comprising 104 videos with paired tympanometry across six diagnostic categories. In this exploratory paired-cohort analysis, the video-only CNN achieved an overall accuracy of 63.46% (95% CI, 53.88–72.08), while the hybrid video-plus-tympanometry framework achieved 81.73% accuracy (95% CI, 73.22–87.98) at the best observed threshold of 0.90, corresponding to an observed absolute improvement of 18.27 percentage points over the CNN-only model (95% paired bootstrap CI, 9.62–26.92). Improvements were observed for selected categories represented in this cohort, including effusion and retraction, although confidence intervals were wide for categories with small positive sample sizes. These findings suggest that integrating anatomical video information with physiological tympanometry may improve AI-assisted middle ear diagnosis in this external paired cohort, but the operating threshold and per-class performance require confirmation in larger prospective paired multimodal cohorts.
Related
- TAPAS: Learned integration of AlphaFold3 confidence and geometric features for TCR-pMHC binding prediction
- Decision Confidence Neuron in Echo State Network for Continual Evaluation of EEG Motor Imagery Classification Quality
- Expert consensus on competency enhancement of pertussis prevention, diagnosis, and management for healthcare professionals through medical-preventive integration.
- Evaluation of the quantiplus rabies virus detection kit for diagnosis of rabies in humans and animals
- Sensory evaluation of β-ionone in Négrette red wine among sensitive and hyposmic young adults: Integration of genotyping, descriptive, and hedonic analyses
BibTeX
@article{lu2026confidenceguided,
title = {{Confidence-guided integration of otoscopy videos and tympanometry for middle ear disease diagnosis: A multi-center otoscopy study with external paired-cohort evaluation}},
author = {Hao Lu and Carl D. Langefeld and Amy Zinnia and Muhammet Fatih Demir and Shalaka Chavan and Charles A. Elmaraghy and Margaret Wilson and Muhammad Khalid Khan Niazi and Aaron C. Moberly and Metin N. Gurcan},
journal = {PLOS One},
year = {2026},
month = sep,
doi = {10.1371/journal.pone.0357492},
url = {https://doi.org/10.1371/journal.pone.0357492},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Confidence-guided integration of otoscopy videos and tympanometry for middle ear disease diagnosis: A multi-center otoscopy study with external paired-cohort evaluation
AU - Hao Lu
AU - Carl D. Langefeld
AU - Amy Zinnia
AU - Muhammet Fatih Demir
AU - Shalaka Chavan
AU - Charles A. Elmaraghy
AU - Margaret Wilson
AU - Muhammad Khalid Khan Niazi
AU - Aaron C. Moberly
AU - Metin N. Gurcan
PY - 2026
DA - 2026/09/01
JO - PLOS One
DO - 10.1371/journal.pone.0357492
UR - https://doi.org/10.1371/journal.pone.0357492
AB - Accurate diagnosis of middle ear diseases using otoscopy remains challenging, particularly in primary care settings where clinician experience with otoscopy varies and visual examination alone may not fully capture middle ear physiology. Although AI-based models have shown promise for automated otoscopy interpretation, most rely primarily on visual information, creating an opportunity to improve diagnostic robustness by integrating complementary physiological measurements. Here, we present OtoTymp-AI, a confidence-guided multimodal decision-fusion framework that integrates otoscopy video analysis with conventional 226 Hz tympanometry. A convolutional neural network was trained to classify otoscopy videos, and tympanometric measurements were incorporated through a clinically interpretable, rule-guided decision strategy to refine visually uncertain predictions rather than through jointly trained multimodal representation learning. Within this multi-center otoscopy study, multimodal evaluation was performed in one independent external cohort comprising 104 videos with paired tympanometry across six diagnostic categories. In this exploratory paired-cohort analysis, the video-only CNN achieved an overall accuracy of 63.46% (95% CI, 53.88–72.08), while the hybrid video-plus-tympanometry framework achieved 81.73% accuracy (95% CI, 73.22–87.98) at the best observed threshold of 0.90, corresponding to an observed absolute improvement of 18.27 percentage points over the CNN-only model (95% paired bootstrap CI, 9.62–26.92). Improvements were observed for selected categories represented in this cohort, including effusion and retraction, although confidence intervals were wide for categories with small positive sample sizes. These findings suggest that integrating anatomical video information with physiological tympanometry may improve AI-assisted middle ear diagnosis in this external paired cohort, but the operating threshold and per-class performance require confirmation in larger prospective paired multimodal cohorts.
SN - 1932-6203
ER -