A novel noninvasive approach for evaluating work of breathing indices in a developmental rat model using respiratory inductance plethysmography

Pulmonary function testing (PFT) is an important component for evaluating the outcome of experimental rodent models of respiratory diseases. Respiratory inductance plethysmography (RIP) provides a noninvasive method of PFT requiring minimal cooperation. RIP measures work of breathing (WOB) indices including phase angle (Ф), percent rib cage (RC %), breaths per minute (BPM), and labored breathing index (LBI) on an iPad. The aim of this study was to evaluate the utility of a recently developed research instrument, pneuRIP, for evaluation of WOB indices in a developmental rat model. Sprague Dawley rats (2 months old) were commercially acquired and anaesthetised with isoflurane. The pneuRIP system uses two elastic bands: one band (RC) placed around the rib cage under the upper armpit and another band (AB) around the abdomen. The typical thoracoabdominal motion (TAM) plot showed the abdomen and rib cage motion in synchrony. The plots of phase angle and LBI as a function of data point number showed that values were within the range. The distribution for phase angle and LBI was within a narrow range. pneuRIP testing provided instantaneous PFT results. This study demonstrated the utility of RIP as a rapid, noninvasive approach for evaluating treatment interventions in the rodent model.


Scientific Reports
| (2020) 10:20730 | https://doi.org/10.1038/s41598-020-77731-6 www.nature.com/scientificreports/ non-anaesthetised rodents to evaluate breathing pattern and to detect sensory irritation, pulmonary irritation, and airflow limitation 7,8 . Respiratory inductance plethysmography (RIP) and pneumotachography have been developed to measure pulmonary function in paediatric patients who are minimally cooperative 9,10 . The RIP technique uses the Respitrace system (Sensormedics, Yorba Linda, CA, USA), a cart-mounted, wired system 11 ; breathing pattern and only a few parameters can be viewed on the monitor during the test, and the data analysis has to be processed off-line and is time consuming, which limits timely diagnosis and treatment decisions [11][12][13] . In a rat model of native airway hyper-responsiveness, at the endpoint of the study, invasive pulmonary function measurement was conducted to measure respiratory resistance and compliance on tracheostomised and mechanically ventilated mice using the plethysmography system of the plethysmography/ventilator chamber FinePointe RC (Buxco Research Systems, Wilmington, NC, USA) 14 .
Recently, our group has utilized the pneuRIP system (Creative Micro Designs, Newark, DE, USA), a new, noninvasive pulmonary function instrument that employs the RIP technique to test infants and children with pulmonary disorders. The hardware of this system includes a palm-sized hardware module and an iPad (Apple, Cupertino, CA, USA). The pneuRIP can wirelessly receive RIP data from the rodent and displays real-time work of breathing (WOB) parameters instantaneously on the iPad 15 . The WOB indices include phase angle (Ф), percent rib cage (RC %), breaths per minute (BPM), and labored breathing index (LBI). Benchtop experiments have been conducted to test the reliability, sensitivity, and reproducibility of this system 15 . Furthermore, the pneuRIP system was tested with 10 normal children in normal breathing conditions and with an added resistive load to compare with an existing system, Respitrace. The two systems were consistent in the measurement of phase angle, RC %, and LBI. The pneuRIP is more accurate than the Respitrace system for breathing rate 15 . The pneuRIP system and Respitrace are interchangeable in measuring the WOB indices. In a recent clinical study, our group performed pneuRIP testing on 9 healthy volunteers and 7 patients with Duchenne muscular dystrophy (DMD). The patients with DMD showed diffuse breathing patterns and a significant elevation in Φ, LBI, and heart rate (HR). Healthy subjects and patients with DMD had similar BPM and RC % averages. A mean ± standard error of the mean (SEM) of phase angle of 23.40° ± 4.33° and LBI of 1.07 ± 0.02 were reported in healthy subjects 16 .
While the normal values of pulmonary function tests in humans who are conscious and cooperative are well established, little is known regarding pulmonary function values such as WOB indices in anaesthetised rodents. This study is the first to use this newly developed RIP technique, pneuRIP, on rodents. As such, the aim of this feasibility study was to evaluate the utility of the developed research instrument, pneuRIP, for evaluation of WOB indices in a developmental rat model.

Materials and methods
Animal preparation. This study was approved by the Institutional Animal Care and Use Committee (IACUC) of Nemours/Alfred I. duPont Hospital for Children, and all methods were performed in accordance with the National Institutes of Health's Guide for the Care and Use of Laboratory Animals. This pilot study was conducted with 5 Sprague Dawley rats (245.9 ± 55.8 g, 2 months old) that were commercially acquired (Charles River Laboratories, Malvern, PA, USA). Rats were anaesthetised for 3 min with an induction dose of 5% isoflurane vapor with 40% fraction of inspired oxygen (FiO 2 ) and maintained with 3% inhaled isoflurane with 40% FiO 2 . The rats were placed in a right lateral lying position on an infrared warming pad (Kent Scientific Corporation, Torrington, CT, USA) during 3.5 min of testing. Oxygen saturation and heart rate were monitored using a pulse oximeter (MouseSTAT, Smiths Medical, Waukesha, WI, USA). End-tidal carbon dioxide (ETCO 2 ) was measured using a Non-invasive Cardiac Output (NICO) monitor (Novametrix Medical Systems, Wallingford, CT, USA). pneuRIP measurement. Respiratory inductance plethysmography using the pneuRIP (hardware and software provided by Creative Micro Designs, Newark, DE, USA) was conducted on the 5 Sprague Dawley rats. The pneuRIP system uses two elastic bands embedded with insulated induction wires, which measure the movement and synchrony of the rat's rib cage and diaphragm 17 ; one band (RC) is placed around the rib cage under the armpit of the upper limb and another band (AB) is placed around the abdomen at the largest abdominal circumference parallel to the RC band (Fig. 1a). www.nature.com/scientificreports/ The pneuRIP wirelessly connects with the iPad to plot the data points and calculates the RIP parameters (Fig. 1b). During 3.5 min of testing, movements of the rib cage and abdomen were plotted, and WOB indices were calculated and continually recorded in real time, which included LBI, Ф between abdomen and rib cage, respiratory rate, and RC % 15,16 .
As previously described 15,16 , WOB indices are defined as follows: (1) Percent rib cage (RC %), which is a ratio of the absolute magnitude of RC signal to the sum of the absolute magnitudes of the RC and ABD signals. The RC %, shown in Eq. (1), defines RC contribution to tidal volume and is presented as a percent of the combined RC and ABD volume changes.
where |RC| is the magnitude of the signal recorded from the rib cage and |ABD| is the magnitude of the signal recorded from the abdominal compartment. (2) Labored breathing index (LBI) (Eq. 2), which is a measure of respiratory effort due to asynchronous breathing. LBI is a ratio that is an estimate of WOB effort.
where |RC| t and |ABD| t are the magnitudes of the RC and ABD compartment signals at any instant of time (t), while |RC| where x represents RC, y represents ABD, and n is the number of samples. (4) Respiratory rate in breaths per minute. This was determined by a fast Fourier transform (FFT) algorithm to generate the average magnitudes for each frequency of the RC and ABD signal. From these calculations, an array is generated from which the largest value is selected, as well as amplitudes within 20% of this value. Finally, we compute a weighted average using indices of these values. This value is the frequency multiplied by 60 to obtain breaths per minute.
During the RIP measurements, patterns of breathing were determined by recording the signals of the rib cage and abdomen. After the test was completed, the data recordings were securely saved on the iPad and were transferred via secured email for further analysis. Breaths per minute (BPM) were calculated at a rate of 10 measurements per second.
Statistical analysis. Statistical software SPSS version 25 (IBM, Armonk, NY, USA; https ://www.ibm.com/ produ cts/spss-stati stics ) and Excel 2016 (Microsoft, Redmond, WA, USA) were used to perform the statistical analysis and data plotting. The pulmonary function summary data were presented in tabular form as mean and SEM (Table 1). To date, there have been little data reported on RIP in rats, so we compared our rat data with other RIP data from human infants and children. Typical plots of Ф and LBI were made as a function of data point number (10 points per second) in an individual rat and data collection period. In addition, the overall distribution histograms for all five rats were plotted to identify the distribution of group breathing patterns over time and for comparison with previous studies. www.nature.com/scientificreports/

Results
The mean and SEM of physiological and pulmonary function parameters are summarized in  Figure 2 shows a typical plot of the thoracoabdominal motion (TAM) as a function of sample number/time in the same rat and time segment. As illustrated, the abdomen and rib cage motion were of similar amplitude and almost superimposed as a function of sample number/time, thus indicating a synchronous breathing pattern consistent with a low mean phase angle of 4.42° and mean LBI of 1.001.
With regard to WOB indices, most noteworthy are the means and distribution of phase angle and LBI over the study duration 11,12,15,16 . As shown in Fig. 3, typical measurements of phase angle and LBI of the same rat and time segment are plotted as a function of data point number/time over the study duration. As illustrated in Fig. 3a, the phase angle values were within the range (0 ≤ Ф ≤ 20°). As shown in Fig. 3b, the LBI values were within the range (1 ≤ Ф ≤ 1.16).
Phase angle and LBI group histograms of all 5 rats are plotted in Fig. 3c and d to demonstrate how these RIP parameters are distributed as a function of measurement range. The phase angle histogram (Fig. 3c) shows a clustering of measured data points below 30° and demonstrates that 96.2% of the phase angle data values were within the range (0 ≤ Ф ≤ 30°). As shown in Fig. 3d, the LBI histogram shows a clustering of measured data points between 1 and 1.10; the LBI histogram demonstrates that 88.5% of the data values were within the range (1.00 < LBI < 1.10).

Discussion
In this study, noninvasive pneuRIP testing was performed on anaesthetised, spontaneously breathing, 2-monthold rats. The primary objective was to demonstrate the utility of the newly developed pneuRIP system for evaluation of WOB indices including Ф, RC %, BPM, and LBI for developing rats.
The relationship between age of laboratory rats and human has been established 1,2 . Rats become sexually mature at about 6 weeks old. In adulthood, every month of the animal is approximately equivalent to 2.5 human years 1 . The rats used for the current study were 2 months old, which is equivalent to young adult in humans 2 . In most cases, the PFT changes in small animals correspond to changes in human PFTs 4 . In general, for humans and most mammals, normal synchronous breathing is observed at Ф < 30° and LBI < 1.05 18 . A phase angle of 0° represents perfect synchrony between the chest and abdominal compartments 19 . Higher phase angles represent greater degrees of asynchronous breathing, and a phase angle of 180° represents paradoxical breathing 13 . The LBI is an indicator of the additional WOB required to maintain gas exchange when breathing is asynchronous and is associated with pulmonary dysfunction [10][11][12]16,17 .
As shown in Table 1, the mean ± SEM of phase angle of 8.94 ± 0.22° and LBI of 1.04 ± 0.001 for rats were considered within the normal range compared with humans based on our previous studies 15,17,18 as well as others [10][11][12] . The results of HR, SPO 2 , ETCO 2 , and BPM were consistent with the results in our previous rat study 14 .
As shown in Fig. 2, the typical TAM plot showed the abdomen and rib cage motion both in synchrony, which was similar to the pattern of real-time plots displayed on the iPad in previous healthy human studies [10][11][12]15,16 . Figure 3a and b illustrate typical plots of phase angle and LBI as a function of data point number and show that the values of phase angle and LBI were within the range (0 ≤ Ф ≤ 20°, 1 ≤ LBI ≤ 1.16, respectively). These data demonstrated the study rats had synchronous breathing and, coupled with normal cardiopulmonary measurements (Table 1), indicated no pulmonary dysfunction. As shown in Fig. 3c and d, data from all five rats were plotted as histograms indicating data distribution over the study duration. The phase angle histogram (Fig. 3c) showed a clustering of measured data points below 30° and demonstrated that 96.2% of the phase angle data values were within the range (0 ≤ Ф ≤ 30°) as shown for healthy humans 15,18 . In Fig. 3d, the LBI histogram showed a clustering of measured data points between 1 and 1.10; the LBI histogram demonstrated that 88.5% of the www.nature.com/scientificreports/ data values were within the normal range. The distribution for phase angle and LBI was within a narrow normal range, a single-distribution clustering 15,18 . These WOB index findings were consistent with those of our previous human studies 12,16,17 . Warren et al. performed a semi-quantitatively calibrated, noninvasive RIP on 61 preterm infants and 7 fullterm infants. Mean (± SD) phase angle was significantly greater in preterm infants than in full-term infants (60.6 ± 39.8 versus 12.5 ± 5.0). Mean (± SD) LBI was significantly greater in preterm infants than in full-term infants (1.35 ± 0.35 versus 1.01 ± 0.01). The rib cage contribution to breathing did not differ significantly between preterm and full-term infants 13 . The phase angle and LBI data of full-term infants support the findings of our current study, in which phase angle and LBI were the key measurements of synchrony related to respiratory function.
Chest wall motion (CWM) was evaluated using RIP before and during mechanical ventilation to evaluate five children and young adults with neuromuscular (NM) disease. It was shown that mechanical ventilation improved RC/ABD asynchrony and decreased Φ in this NM patient population 20 . As shown, the authors concluded that assessment of CWM with RIP may be useful in evaluating respiratory function in an intensive care setting, as well as assisting in the determination of optimal ventilator settings in NM patients.
Over the years, RIP measurements have been utilized in determining differences between healthy subjects and patients with respiratory complications using post-hoc analysis with Respitrace. In this regard, Giordano et al. 10 showed that RIP data correlated with admission criteria in pediatric asthmatic patients in the emergency department. They found that Φ, LBI, and HR were all elevated in those patients who were admitted to the hospital compared with those who were discharged to home. Furthermore, Rodriguez et al. 21 demonstrated that RIP measurements could identify thoracic problems in children with skeletal dysplasia. Another skeletal dysplasia study demonstrated the utility of RIP in Morquio patients in combination with spirometry and forced oscillation techniques 22 . Kubaski et al. noted in a large population of Morquio patients 3 to 40 years of age that these patients had normal breathing synchrony using combined RIP and impulse oscillometry technology. Although these patients had small lung volumes due to Morquio syndrome, they exhibited no clinical or functional signs of restrictive or obstructive lung disease. It is noteworthy that RIP was essential in identifying that patients with Morquio have small lungs with synchronized breathing while maintaining normal gas exchange at rest 22 .

Study limitations.
Noninvasive pneuRIP testing provides instantaneous PFT results requiring minimal patient cooperation as previously noted above. However, rodents need to be anaesthetised with either inhalant gas or injectable drugs during the measurement process. The use of inhalant gas is the preferred method www.nature.com/scientificreports/ of anaesthesia whenever possible. During testing, it is important to control the depth of anaesthesia and keep constant positioning across measurements and the same anaesthesia level for each rat. If rats are exposed to excessive isoflurane, the rats' breathing would become shallow and irregular, and some signals of the TAM would become too weak to be detected by the pneuRIP software. In this study, it should be noted that the rat sample size was limited; thus, more normal rats would be necessary to measure to represent the population values for all rats. With regard to repeatability and credibility of the pneuRIP system, it has been validated with respect to bench studies and healthy volunteer comparisons with the gold standard, Respitrace 15 . It is noteworthy to mention that the variations in WOB indices (phase angle and LBI) are quite small and that the normality of phase angle and LBI for the presented rodents is similar to that reported for healthy humans in the age range of 3 to 18 years 15,17,18 . Conclusion. The pneuRIP testing provided instantaneous PFT results for rapid and noninvasive study of rodent breathing patterns and WOB indices. The pneuRIP method was precise and accurate in measuring the phase angle, RC %, and LBI. This novel approach for assessment of PFT translated appropriately to a small experimental rodent model, and, in addition to providing immediate results, the method has the potential for studying longitudinal PFT outcomes in a developing rat model. As previously noted, this experimental model has been used to study many respiratory neonatal disorders such as hyperoxia, barotrauma, sepsis, etc.

Data availability
The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.