Pressure-sensitive mats (PSM) have proved to be useful in the estimation of respiratory rates (RR) in adult patients. However, PSM technology has not been extensively applied to derive physiologic parameters in infant and neonatal patients. This research evaluates the applicability of the capacitive XSensor PSM technology to estimate a range of RR in neonatal patient simulator trials conducted under several experimental conditions. PSM data are analyzed in both the time and frequency domain and comparative results are presented. For the frequency-domain approach, in addition to estimating RR, a measure of confidence is also derived from the relative height of peaks in the periodogram. The study demonstrates that frequency domain analysis of mean-shifted PSM data achieves the best possible RR estimation, with zero percent error, as compared to the lowest achievable RMS error of 1.57 percent in the time domain. The frequency domain approach outperforms the time domain analysis whether examining raw data or those preprocessed by normalizing, detrending and median filtering.
There is increasing evidence that the respiratory rate (RR) of a person lying on a pressure-sensitive mat (PSM) can be estimated by analyzing the measured contact pressure data. Past research that demonstrates the utility of PSM data for the estimation of RR has been mainly focused on the adult population, such as [1]- [12]. Two systematic reviews of noninvasive respiratory monitoring in clinical care, conducted almost a decade apart [13], [14], were similarly focused on adult studies. The most recent review, in 2015, concludes that such monitoring has the potential for improved early diagnosis of patient deterioration and the reduction of critical events for patients on the general wards [14]. PSM technology is wellsuited to long-term patient monitoring both at home and in hospitals due to it being non-invasive, contactless, and unobtrusive. However, when it comes to the infant and neonatal population, there is a scarcity of research within the field.
In 1976, Franks et al. pioneered the idea to measure pressure changes in the thorax region during the respiratory cycle in overnight recordings in infants at home who were up to six months old [15]. They recommended that an undermattress pressure sensor was the most simple and satisfactory amongst five contactless devices tested, especially in cases where there could be considerable parental apprehension. However, the use of PSM has not yet been evaluated on critically ill term and preterm babies within a hospital’s neonatal intensive care unit (NICU). Previously, the use of mechanical sensors such as single-point load cells [16], pressure sensors [17], [18], and piezoceramic sensors [19] has been researched for the extraction of neonatal and infantile RR. The key disadvantage with all these methods is the loss of spatial resolution when compared to PSM technology, where an array of pressure transducers provides a time-varying highdensity 2D map of contact pressure. Secondly, sensors that attach to the patient’s bed are relatively difficult to transfer should the patient needs to be transferred to another bed.
A number of infant breathing monitors that use mechanical sensors are available in the market including wearables such as Snuza Hero, Levana Oma, Mimo and BabySense [20]. However, these systems are neither licensed as medical devices for use in hospitals in North America nor do they provide application program interfaces for secondary analyses of their datasets.
As we begin to explore the use of PSM as a long-term continuous patient monitoring modality in the NICU, there exist a number of challenges. A major difference that exists in using PSM with the neonatal population, as opposed to adults or older children, is the much smaller mass and pressure ranges [16]. It may be necessary to place the PSM above the mattress but below a crib sheet, since output signals below a mattress can be heavily muted and may feature less distinct localized loading thus reducing the signal-to-noise ratio (SNR) [4]. Thereby, adherence to medical device standards and testing for electrical safety become of utmost importance. In addition, electrical, mechanical, and environmental artifacts decrease the ability to estimate the RR accurately. These include power line interference, electrical noise, patient movements, and vibrations from people walking around the bed, possible air currents from heating, ventilation and air conditioning. Physiologic artifacts are another potentially confounding factor in the detection of the RR signal from PSM data. For example, expiratory grunting distorts the breathing signal and produces an unclear noisy expiration wave [21], [22]. Grunting is a compensatory breathing effort made by a neonate to overcome some lung abnormality. Grunting is produced by the expiration of air through partially closed vocal cords, either intermittently or continuously depending on the severity of the lung disease [23]. Clinicians typically hear the grunting sound produced by a patient. Grunting is one of the symptoms of neonatal respiratory distress syndrome (NRD) and is typically recognized by clinicians when listening to a patient breath [24]. Term and preterm infants presenting with NRD are admitted to the NICU and treated with oxygen supply [25], [26]. NRD accounts for significant morbidity and mortality [27].
In this proof-of-concept study, we collect a comprehensive dataset from a capacitive PSM LX100:36.36.02 (XSensor Technology Corp. Calgary, Canada, XSensor.com) to show that it is possible to extract simulated neonatal RR from PSM data. In our previous work, the XSensor PSM technology has shown a superior dynamic response as compared to the metrological properties of resistive Tekscan and optical S4 PSM technologies [28]. This study compares time and frequency domain approaches for RR estimation with results presented in terms of estimation error and confidence.
A breathing pattern is a time series signal dominated by respiratory modulation [29]. To c
This content is AI-processed based on open access ArXiv data.
