Introduction: PhotoniCare, Inc. is a medical device company developing the TOMi Scope, a handheld, optical imaging device for improved diagnosis of middle ear health. The purpose of this proposal is to establish and evaluate a machine learning approach to interpret TOMi Scope depth-resolved images using a set of ear models with human middle ear effusion (MEE; fluid) to enable improved diagnostic accuracy and, ultimately, antibiotic stewardship for ear health. Significance: Ear infections affect 95% of all children, yet they are one of the most poorly diagnosed and managed diseases in all of medicine, resulting in high antibiotic over-prescription and antibiotic resistance development. Correctly identifying the absence or presence/type of MEE through the non-transparent eardrum is critical to accurate diagnosis, and the limited current diagnostic tools suffer poor diagnostic accuracy (50- 70%) due to inherent subjectivity and dependence on user experience. Therefore, objective image classification metrics to enable improved diagnostic accuracy is sorely needed to finally provide children afflicted by this disease with the correct treatment the first time. Hypothesis: Applying a machine learning approach to TOMi Scope image classification of a set of ear models with human MEE will facilitate detection of the presence or absence of effusion (?90% accuracy), as well as classification by the type of effusion samples (?80% accuracy), regardless of user experience.
Specific Aims : (1) Collect robust datasets of ex vivo human MEE, sufficient for machine learning image analysis. (2) Develop a neural network model based on the MEE dataset and apply the model to a representative test clinical dataset to determine classification feasibility. Commercial Opportunity: The TOMi Scope will provide physicians with new, objective information, enabling better decision-making for antibiotic prescription and surgical intervention. This has the potential to impact the standard of care for ~1B children worldwide that experience ear infections, representing a multi-billion-dollar commercial opportunity.
Ear infections (otitis media) are highly prevalent in the pediatric population and represent a significant clinical challenge due to the limitations of the gold-standard diagnostic tools, resulting in high antibiotic prescription but also antibiotic resistance development. Accurate detection and classification of effusion (fluid) in the middle ear is a critical element for this diagnosis, and for making informed medical treatment decisions, particularly regarding antibiotic stewardship. The long-term goal of this work is to reduce antibiotic resistance and healthcare costs through improving patient outcomes by addressing the low diagnostic accuracy and user experience dependence of current subjective methods, with a novel, non-invasive imaging tool capable of quantitative depth-resolved measurements to not only visualize the underlying infection behind the eardrum, but also, with automated machine learning image analysis algorithms, minimize user experience dependence and variability.
