Detecting middle ear fluid using smartphones.

Area of Science:

• Biomedical Engineering
• Otolaryngology
• Machine Learning in Healthcare

Background:

• Middle ear fluid presence is crucial for diagnosing acute otitis media and otitis media with effusion in children.
• Current diagnostic methods often require specialized equipment and trained personnel.

Purpose of the Study:

• To develop and validate a smartphone-based system for detecting middle ear fluid by assessing eardrum mobility.
• To compare the performance of the smartphone system against commercial acoustic reflectometry and established diagnostic tools.

Main Methods:

• A clinical study involving 98 patient ears was conducted at a pediatric surgical center.
• A machine learning algorithm analyzed data from smartphone speakers and microphones to assess eardrum mobility.
• Leave-one-out cross-validation was used to estimate algorithm performance on unseen data.

Main Results:

• The smartphone-based algorithm achieved an Area Under the Curve (AUC) of 0.898, outperforming commercial acoustic reflectometry (AUC = 0.776).
• The system demonstrated 85% sensitivity and 82% specificity, comparable to tympanometry and pneumatic otoscopy.
• Similar performance was observed across multiple smartphone platforms, and parents achieved comparable results to clinicians.

Conclusions:

• A smartphone can serve as a low-barrier, effective screening tool for middle ear fluid detection.
• This technology has the potential to improve accessibility and efficiency in diagnosing pediatric ear conditions.
• Further validation could lead to widespread use in clinical and potentially home settings.