loading page

Artificial intelligence to classify ear disease from otoscopy: A systematic review and meta-analysis
  • +6
  • Al-Rahim Habib,
  • Majid Kajbafzadeh,
  • Zubair Hasan,
  • Eugene Wong,
  • Hasantha Gunasekera,
  • Christopher Perry,
  • Raymond Sacks,
  • Ashnil Kumar,
  • Narinder Singh
Al-Rahim Habib
The University of Sydney

Corresponding Author:[email protected]

Author Profile
Majid Kajbafzadeh
The University of Sydney
Author Profile
Zubair Hasan
Westmead Hospital
Author Profile
Eugene Wong
Westmead Hospital
Author Profile
Hasantha Gunasekera
The University of Sydney
Author Profile
Christopher Perry
Brisbane ENT
Author Profile
Raymond Sacks
Sydney Medical School
Author Profile
Ashnil Kumar
The University of Sydney
Author Profile
Narinder Singh
University of Sydney/ Westmead Hospital
Author Profile

Abstract

Objective: To summarize the accuracy of artificial intelligence (AI) computer vision algorithms to classify ear disease from otoscopy. Methods: Using the PRISMA guidelines, nine online databases were searched for articles that used AI methods (convolutional neural networks, artificial neural networks, support vector machines, decision trees, k-nearest neighbors) to classify otoscopic images. Diagnostic classes of interest: normal tympanic membrane, acute otitis media (AOM), otitis media with effusion (OME), chronic otitis media (COM) with or without perforation, cholesteatoma, and canal obstruction. Main Outcome Measures: Accuracy to correctly classify otoscopic images compared to otolaryngologists (ground-truth). The Quality Assessment of Diagnostic Accuracy Studies Version 2 tool was used to assess the quality of methodology and risk of bias. Results: Thirty-nine articles were included. Algorithms achieved 90.7% (95%CI: 90.1 – 91.3%) accuracy to difference between normal or abnormal otoscopy images in 14 studies. The most common multi-classification algorithm (3 or more diagnostic classes) achieved 97.6% (95%CI: 97.3.- 97.9%) accuracy to differentiate between normal, AOM and OME in 3 studies. Compared to manual classification, AI algorithms outperformed human assessors to classify otoscopy images achieving 93.4% (95%CI: 90.5 – 96.4%) versus 73.2% (95%CI: 67.9 – 78.5%) accuracy in 3 studies. Convolutional neural networks achieved the highest accuracy compared to other classification methods. Conclusion: AI can classify ear disease from otoscopy. A concerted effort is required to establish a comprehensive and reliable otoscopy database for algorithm training. An AI-supported otoscopy system may assist health care workers, trainees, and primary care practitioners with less otology experience identify ear disease.
11 Nov 2021Submitted to Clinical Otolaryngology
17 Nov 2021Submission Checks Completed
17 Nov 2021Assigned to Editor
24 Nov 2021Reviewer(s) Assigned
16 Dec 2021Review(s) Completed, Editorial Evaluation Pending
26 Dec 2021Editorial Decision: Revise Minor
07 Jan 20221st Revision Received
10 Jan 2022Submission Checks Completed
10 Jan 2022Assigned to Editor
24 Jan 2022Reviewer(s) Assigned
23 Feb 2022Review(s) Completed, Editorial Evaluation Pending
27 Feb 2022Editorial Decision: Accept
May 2022Published in Clinical Otolaryngology volume 47 issue 3 on pages 401-413. 10.1111/coa.13925