Continuous Respiratory Monitoring Platform for Classifying Apnea of Prematurity - Summary/Abstract Apnea of prematurity (AOP) is a prevalent condition in premature newborns that can result in both short and long-term complications ranging from cardiorespiratory failure to neurological impairment. Despite its prevalence, existing standards for monitoring AOP are inadequate, which leads to sub-optimal patient treatment and a general lack of understanding of the condition and its best practices. There is a significant need for an accurate and continuous apnea monitor that can better characterize AOP, giving healthcare providers and researchers the tool to optimize patient care, enhance safety, and advance evidence-based practices. The goal of this proposal is to develop an AOP-specific monitor capable of identifying the major apnea types (central obstructive, and mixed) in premature newborns. This work will be performed by Makani Science in collaboration with the Children’s Hospital of Orange County. We propose that apnea can be better identified using a multimodal monitor with sensors specific to AOP and optimized for newborns. When paired with machine learning, we will be able to classify apnea accurately and automatically. The proposed work for this grant will be accomplished in two phases (R61 and R33). During the R61 phase, a multimodal system and machine learning model will be developed and validated for detecting AOP. This phase will have three specific aims: (1) develop hardware and software prototype for monitoring apnea, (2) collect initial AOP data and characterize features for classical machine learning models, and (3) diversity AOP dataset and explore deep learning-based models. During the R33 phase, the prototype will be optimized into a full integrated system that will be ready for design freeze. This phase will also have three specific aims: (1) verify device design and acquire feedback form the FDA, (2) optimize the system and integrate a stimulator for terminating apnea, and (3) clinically validate the integrated system. Successful completion of this proposal will result in an accurate apnea monitor that can enhance patient safety and reduce workload for healthcare professionals. The monitor will also give researchers an accessible tool for collecting data on AOP. Additionally, this effort itself will create a comprehensive database that can help advance our understanding of the condition.
