Fetal Monitoring Signal Optimization Device

Abstract

This project seeks to improve the safety and reliability of fetal and maternal monitoring during labor and delivery by optimizing how physiologic signals are processed and interpreted. Our team is developing a Hardware-Based Filtering & Optimization/Machine Learning (HBFO/ML) module, a plug-in accessory that is designed to be attached to already existing fetal/maternal monitors. It is an intermediate step between the existing device’s data collection and its digitally processed output, which allows for the generation of a cleaner biosignal through hardware filtering and subsequently improved signal interpretations through the use of machine learning algorithms for better event classification.

The goal of the system is to significantly reduce false alarms during decelerations and contractions and to improve signal discrimination between fetal heart rate, maternal heart rate, and uterine activity. Current fetal/maternal monitoring devices have a high rate of false positives because the collected biosignals are weak and prone to noise, making accurate interpretation challenging. By addressing these shortcomings through principles of digital signal processing, embedded machine learning, and closed-loop optimization, the module aims to reduce adverse fetal and maternal outcomes, such as unnecessary cesarean delivery and perinatal mortality. Our device will achieve this goal through a combination of active filters and amplifiers for the hardware-based filtering, and a sophisticated machine learning module, which will use publicly available clinical datasets in conjunction with personalized patient data to analyze and predict fetal and maternal health conditions in real time.

This module will hold most of its costs within the design and upkeep of the machine learning software, where we will require a data center to securely keep all sensitive patient data. The intended commercialization pathway is B2B: the module is designed to be distributed to companies that manufacture fetal and maternal monitors as a performance-enhancing accessory rather than a replacement device.

Video