Untitled design(4)

App captures eardrum videos and expert annotated descriptions with handheld, otoscope/iPhone device, voice recognition to start/stop video capture.

Ear Infections are Over-Diagnosed and Antibiotics Over-Prescribed

The Pitt-CMU iTM Technology Automatically Diagnoses Acute Otitis Media (AOM) and More Than $4B are spent for AOM.

SUMMARY:

  • More than 20M prescriptions per year = $2B direct and $2B indirect costs
  • Up to 50% of ear infection diagnoses are wrong
  • Only 75% accuracy with experienced clinicians

There is a need for simple, reliable clinical decision support to improve AOM diagnosis and reduce unnecessary antimicrobial therapy.

Automated Diagnosis of Acute Otitis Media

  • Handheld, otoscope/iPhone device diagnoses otitis media by classifying tympanic membrane images
  • App captures eardrum videos and expert annotated descriptions
  • IP-protected 8-feature algorithmic AI with built-in noisy frame rejection AI
3-May-18-2023-02-49-46-8039-PM

Advantages

Technology Graphic 3-1

Proof of concept

Initial validation of in 200 eardrum images – 90% accuracy

AI development in 1150 videos with expert annotations– 94% accuracy

Technology Graphic 1-1

Meaningful Features

Targeted, physiologically meaningful features yield higher accuracy vs. general features (90% vs. 70%)
Technology Graphic 2-1

Patent Protection

US9636007B2 - Method and apparatus for aiding in the diagnosis of otitis media by classifying tympanic membrane images (View on Google Patents)

US10182711B2 - Method and apparatus for aiding in the diagnosis of otitis media by classifying tympanic membrane images (View on Google Patents)

1-May-18-2023-02-48-55-7968-PM

Meet the Pitt Innovator

Alejandro Hoberman, MD, Executive Vice Chair of Pediatrics, Vice Chair of Clinical Research, University of Pittsburgh