Validating a Deep Learning Algorithm in Children With Ear Concerns
Trial Snapshot
- Phase
- Not Applicable
- Status
- Not yet recruiting
- Sponsor
- Enrollment
- 658
- Primary Endpoint
- Percent agreement of Glimpse machine learning algorithm's classification of a child's ear image with an ENT panel diagnosis
Study Overview
Brief Summary
The goal of this observational study is to determine if the Glimpse machine learning algorithm can accurately assess ear diseases in children. Participants will:
- Have a video of their ear taken by their parent or their guardian
- Have a video of their ear taken by a Primary Care Physician (PCP)
- Have an assessment of their eardrums and a video of their ears taken by an Ear, Nose, and Throat specialist (ENT).
The videos will be used to determine if the Glimpse algorithm matches the diagnosis of the physicians.
Detailed Description
Ear complaints, including earache (otalgia), are the most common reasons children seek healthcare and routinely bring children into the office of a pediatrician or urgent care setting. This study will assess children who present with signs and symptoms of otitis media to the primary care office or urgent care. Participants will receive their standard of care from their treating physician, with study assessments including videos of their ears taken by their parent or guardian and the treating physician. Once this is complete, participants will see an ENT for an assessment of their eardrum. The ENT assessment will occur within 24 hours of the PCP visit and will not be used to inform patient treatment.
Study Design
- Study Type
- Observational
- Observational Model
- Cohort
- Time Perspective
- Prospective
Eligibility Criteria
- Ages
- 6 Months to 6 Years (Child)
- Sex
- All
- Accepts Healthy Volunteers
- No
Inclusion Criteria
- •Males and females aged 6 months to 6 years
- •Presenting to a pediatrician's office or urgent care with signs and symptoms of otitis media, including tugging at ears, ear pain, crying at night, refusing to lie flat, sleeping poorly, having a fever, having decreased appetite, and/or concern for hearing loss, regardless of previous diagnosis of AOM or OME.
Exclusion Criteria
- •History of craniofacial abnormality
- •PE tubes currently in place
- •Current otorrhea
- •Caretaker not having use of both hands and arms
Outcomes
Primary Outcomes
Percent agreement of Glimpse machine learning algorithm's classification of a child's ear image with an ENT panel diagnosis
Time Frame: Within 24 hrs of presenting to PCP or urgent care office
The primary endpoint of this study is to compare the percent agreement of Glimpse machine learning algorithm's classification of a child's ear image with an ENT panel diagnosis of the same child's ear for the diagnoses of acute otitis media (AOM), otitis media with effusion (OME), and no middle ear effusion, versus the percent agreement of primary care provider's (PCP) diagnosis with an ENT panel diagnosis, of in children with otalgia.
Secondary Outcomes
No secondary outcomes reported