Eardrum Exams Take a Digital Turn - Summary - MDSpire
From the Journals

Eardrum Exams Take a Digital Turn

Machine learning analyzes smartphone otoscope images to differentiate between otitis media with effusion and normal tympanic membrane images.

Share

A supervised machine-learning model, developed to analyze smartphone-captured tympanic membrane images, demonstrated high accuracy in detecting middle ear effusion among pediatric patients. Conducted as a case-control study at a tertiary center, the research involved 111 tympanic membrane images evaluated by two otolaryngologists. The model achieved 96% sensitivity and 89% overall accuracy during training, with slightly reduced results in testing. Limitations included small sample size and lack of external validation, as noted by lead author Mohammed K. Alnoury from King Abdulaziz University.

Original Source(s)

Eardrum Exams Take a Digital Turn

  • by Kathryn Wighton

  • April 10, 2026

Related Content

Blood Lead, Cadmium Tied to Higher Tinnitus Risk

Large NHANES analysis links higher blood lead and cadmium levels with tinnitus prevalence in US adults.

PRP Linked to Less Pain, Better Healing Following Tonsillectomy

Randomized trial finds reduced postoperative pain and improved mucosal healing with intraoperative application

Postzoster Neuralgia Linked to Tinnitus

Nationwide cohort analysis finds higher tinnitus incidence during long-term follow-up among patients with postzoster neuralgia, particularly older patients and certain comorbidities.