A Closed Loop Feature Detection Platform for Automated Neonate Cardio‐Respiratory Measurements

Abstract

Sudden Infant Death Syndrome (SIDS, a leading cause of death for neonate infants 1 month to 1 year old) and some forms of neonate death (ND, the leading cause of death for newborns 28 days old or less) are thought to result, in part, from unseen brain abnormalities affecting cardio-respiratory function, but for which no clear genetic or environmental mechanisms are known. Mouse genetic and exposure models present an opportunity to uncover potential molecular and environmental mechanisms in SIDS and ND. However, measuring cardio-respiratory function to model SIDS and ND in neonate mice is expensive, difficult, and inefficient. Previously, we have developed for in-lab facile production an inexpensive face-mask pneumotachograph system for neonate respiratory measurements. A remaining challenge is the time and effort needed to carryout neonate respiratory measurements. For example, the neonate autoresuscitation assay consisting of repeated anoxic exposures followed by recovery requires the full attention of a single observer for the duration of the assay (2-3hrs) for a single subject, thus limiting the number of pups in a litter that can be assayed. To address these inefficiencies, we sought to develop a closed loop feature detection platform for automated neonate cardio-respiratory measurements that can be deployed in large numbers for high throughput parallel studies by a single person. The platform design consists of three sub-systems. First, we leveraged our previously published pneumotachograph face-mask system for precise respiratory measurements (PMID: 28213294). Second, we developed a micro-controller automated bell-housing gas switching system for near square wave induction of respiratory challenges. Third we developed a micro-computer-based data-acquisition system with real-time feature detection and response outputs for controlling gas exposure times. The gas switching system consists of linear and rotational actuated carousel that holds four bell housings into which distinct gas mixes can flow by solenoid valves. Through the rotational and linear actuation driven by a pair of stepper motors, the required bell housing is rotated into position and moved over the end of the pneumotach and challenge (e.g. anoxic) gas flow is initiated by a solenoid valve, all under the control of the microcontroller. Micro-controller action is directed by serial input from a micro-computer receiving cardio-respiratory waveform data through an analog to digital converter board. On the micro-computer, a python program detects specific wave form features in real time (lag = 100ms), such as an apnea and bradycardia. Upon apnea detection, the micro-controller is directed to switch to a rescue gas by retraction of the bell housing, rotation, and extension of a different bell housing containing the rescue gas mix. The resumption of normal breathing and heart rate can also be detected and incorporated into the criteria for the automated initiation of the next anoxic exposure trial. Data is stored for later offline automated analysis. The described apparatus offers an inexpensive approach for automated high throughput neonate cardio-respiratory assessment for facile screening to yield important clues in developmental pathophysiology.