Breath-by-breath comparison of a novel percutaneous phrenic nerve stimulation approach with mechanical ventilation in juvenile pigs: a pilot study.

About one in three critically ill patients requires mechanical ventilation (MV). Prolonged MV, however, results in diaphragmatic weakness, which itself is associated with delayed weaning and increased mortality. Inducing active diaphragmatic contraction via electrical phrenic nerve stimulation (PNS) not only provides the potential to reduce diaphragmatic muscular atrophy but also generates physiological-like ventilation and therefore offers a promising alternative to MV. Reasons why PNS is not yet used in critical care medicine are high procedural invasiveness, insufficient evidence, and lack of side-by-side comparison to MV. This study aims to establish a minimal-invasive percutaneous, bilateral electrode placement approach for sole PNS breathing and thereby enable, for the first time, a breath-by-breath comparison to MV. Six juvenile German Landrace pigs received general anesthesia and orotracheal intubation. Following the novel ultrasound-guided, landmark-based, 4-step approach, two echogenic needles per phrenic nerve were successfully placed. Stimulation effectiveness was evaluated measuring tidal volume, diaphragmatic thickening and tomographic electrical impedance in a breath-by-breath comparison to MV. Following sufficient bilateral phrenic nerve stimulation in all pigs, PNS breaths showed a 2.2-fold increase in diaphragmatic thickening. It induced tidal volumes in the lung-protective range by negative pressure inspiration and improved dorso-caudal regional ventilation in contrast to MV. Our study demonstrated the feasibility of a novel ultrasound-guided, percutaneous phrenic nerve stimulation approach, which generated sufficient tidal volumes and showed more resemblance to physiological breathing than MV in a breath-by-breath comparison.

A physiological model of phrenic nerve excitation by electrical stimulation.

Mechanical ventilation is essential in intensive care treatment but leads to diaphragmatic atrophy, which in turn contributes to prolonged weaning and increased mortality. One approach to prevent diaphragmatic atrophy while achieving pulmonary ventilation is electrical stimulation of the phrenic nerve. To automize phrenic nerve stimulation resulting in lung protective tidal volumes with lowest possible currents, mathematical models are required. Nerve stimulation models are often complex, so many parameters have to be identified prior to implementation. This paper presents a novel, simplified approach to model phrenic nerve excitation to obtain an individualized patient model using a few data points. The latter is based on the idea that nerve fibers are excited when the electric field exceeds a threshold. The effect of the geometry parameter on the model output was analyzed, and the model was validated with measurement data from a pig trial (RMSE in between 0.44 × 10-2and 1.64 × 10-2for parameterized models). The modeled phrenic nerve excitation behaved similarly to the measured tidal volumes, and thus could be used to develop automated phrenic nerve stimulation systems for lung protective ventilation.

Hardware-in-the-loop simulation of vascular cannula interaction.

Traditionally, the functional development, testing and verification of higher level automation, control and safety functions in extracorporeal circulations systems relies very much on the conduction of animal experiments. This especially applies to functions requiring interaction with the patients physiology. Besides ethical aspects, animal experiments have the disadvantage of higher complexity, limited reproducibility and certain pathological conditions can only be represented with difficulty. An alternative to this can be Hardware-in-the-loop (HIL) simulators.In this paper we present a HIL solution to simulate the hydraulic interaction between the cannula and the connected blood vessel. We derive a hydraulic model for an elastic vessel with particular emphasis on negative transmural pressure. In this case the resistance is mainly determined by collapse phenomena. The next section describes the design of an universal resistance actuator that can simulate vascular resistances in the expected range. Combined in the HIL simulator, the simulation model then generates the setpoint for the actuator while simultaneously receiving the resulting internal states of the hydraulic interface. This creates a truly interactive HIL simulator where the device under test interacts in the same way as with a physiological system.

Evaluation of electric phrenic nerve stimulation patterns for mechanical ventilation: a pilot study.

Diaphragm atrophy is a common side effect of mechanical ventilation and results in prolonged weaning. Electric phrenic nerve stimulation presents a possibility to avoid diaphragm atrophy by keeping the diaphragm conditioned in sedated patients. There is a need of further investigation on how to set stimulation parameters to achieve sufficient ventilation. A prototype system is presented with a systematic evaluation for stimulation pattern adjustments. The main indicator for efficient stimulation was the tidal volume. The evaluation was performed in two pig models. As a major finding, the results for biphasic pulses were more consistent than for alternating pulses. The tidal volume increased for a range of pulse frequency and pulse width until reaching a plateau at 80-120 Hz and 0.15 ms. Furthermore, the generated tidal volume and the stimulation pulse frequency were significantly correlated (0.42-0.84, [Formula: see text]). The results show which stimulation parameter combinations generate the highest tidal volume. We established a guideline on how to set stimulation parameters. The guideline is helpful for future clinical applications of phrenic nerve stimulation.

SOLVe: a closed-loop system focused on protective mechanical ventilation.

BACKGROUND: Mechanical ventilation is an essential component in the treatment of patients with acute respiratory distress syndrome. Prompt adaptation of the settings of a ventilator to the variable needs of patients is essential to ensure personalised and protective ventilation. Still, it is challenging and time-consuming for the therapist at the bedside. In addition, general implementation barriers hinder the timely incorporation of new evidence from clinical studies into routine clinical practice. RESULTS: We present a system combing clinical evidence and expert knowledge within a physiological closed-loop control structure for mechanical ventilation. The system includes multiple controllers to support adequate gas exchange while adhering to multiple evidence-based components of lung protective ventilation. We performed a pilot study on three animals with an induced ARDS. The system achieved a time-in-target of over 75 % for all targets and avoided any critical phases of low oxygen saturation, despite provoked disturbances such as disconnections from the ventilator and positional changes of the subject. CONCLUSIONS: The presented system can provide personalised and lung-protective ventilation and reduce clinician workload in clinical practice.

A Portable Multi-Modal Cushion for Continuous Monitoring of a Driver's Vital Signs.

With higher levels of automation in vehicles, the need for robust driver monitoring systems increases, since it must be ensured that the driver can intervene at any moment. Drowsiness, stress and alcohol are still the main sources of driver distraction. However, physiological problems such as heart attacks and strokes also exhibit a significant risk for driver safety, especially with respect to the ageing population. In this paper, a portable cushion with four sensor units with multiple measurement modalities is presented. Capacitive electrocardiography, reflective photophlethysmography, magnetic induction measurement and seismocardiography are performed with the embedded sensors. The device can monitor the heart and respiratory rates of a vehicle driver. The promising results of the first proof-of-concept study with twenty participants in a driving simulator not only demonstrate the accuracy of the heart (above 70% of medical-grade heart rate estimations according to IEC 60601-2-27) and respiratory rate measurements (around 30% with errors below 2 BPM), but also that the cushion might be useful to monitor morphological changes in the capacitive electrocardiogram in some cases. The measurements can potentially be used to detect drowsiness and stress and thus the fitness of the driver, since heart rate variability and breathing rate variability can be captured. They are also useful for the early prediction of cardiovascular diseases, one of the main reasons for premature death. The data are publicly available in the UnoVis dataset.

Modelling and Validation of a Decentralized Breathing Gas Source.

In this contribution, the development and validation of a digital twin of a decentralized source of breathing gas used for mechanical ventilation is described. A hardware setup is built in analogy to a Simscape ® model with almost identical architecture and geometry. An experimental trial of volume controlled ventilation was conducted and the obtained data of resulting pressure and oxygen concentration was compared. Correlating measured and simulated data lead to coefficients of 0.83 and 0.74 for FiO2 in the dosage chamber and airway, respectively. Clinical relevance- This work showed that the process of mechanical ventilation can be simulated even to the extent of oxygen distribution for ex-vivo applications.

Identification of the Tidal Volume Response to Pulse Amplitudes of Phrenic Nerve Stimulation Using Gaussian Process Regression.

While mechanical ventilation (MV) can lead to ventilator-induced diaphragmatic atrophy due to diaphragm inactivity, electrical phrenic nerve stimulation (PNS) can keep the diaphragm active and therefore prevent diaphragmatic weakness. To quantify the effectivity of PNS, an identification experiment during PNS is presented, and its data is used in Gaussian process regression (GPR) of the tidal volume based on the constant voltage amplitude of the stimulation pulses. The measurements were split into training data of variable size and test data for cross validation. For variable training sizes and different PNS settings, the GPR had a root mean square deviation (RMSD) between 0.39 and 0.91 mL/kg. An identification experiment as short as one and a half minutes was able to characteristically capture the relationship between tidal volume and voltage amplitude. The proposed method needs to be validated in further experiments.

Fuzzy-Based Expert Supervision System for Feedback Controlled Oxygenation.

Supervision of mechanical ventilation is currently still performed by clinical staff. With the increasing level of automation in the intensive care unit, automatic supervision is becoming necessary. We present a fuzzy-based expert supervision system applicable to automatic feedback control of oxygenation. An adaptive fuzzy limit checking and trend detection algorithm was implemented. A knowledge-based fuzzy logic system combines these outputs into a final score, which subsequently triggers alarms if a critical event is registered. The system was evaluated against annotated experimental data. An accuracy of 83 percent and a precision of 95 percent were achieved. The automatic detection of critical events during feedback control of oxygenation provides an additional layer of safety and assists in alerting clinicians in the case of abnormal behavior of the system. Clinical relevance - Automatic supervision is a necessary feature of physiological feedback systems to make them safer and more reliable in the future.

Comparison of the Hemocompatibility of an Axial and a Centrifugal Left Ventricular Assist Device in an In Vitro Test Circuit.

BACKGROUND: Hemocompatibility of left ventricular assist devices is essential for preventing adverse events. In this study, we compared the hemocompatibility of an axial-flow (Sputnik) to a centrifugal-flow (HeartMate 3) pump. METHODS: Both pumps were integrated into identical in vitro test circuits, each filled with 75 mL heparinized human blood of the same donor. During each experiment (n = 7), the pumps were operated with equal flow for six hours. Blood sampling and analysis were performed on a regular schedule. The analytes were indicators of hemolysis, coagulation activation, platelet count and activation, as well as extracellular vesicles. RESULTS: Sputnik induced higher hemolysis compared to the HeartMate 3 after 360 min. Furthermore, platelet activation was higher for Sputnik after 120 min onward. In the HeartMate 3 circuit, the platelet count was reduced within the first hour. Furthermore, Sputnik triggered a more pronounced increase in extracellular vesicles, a potential trigger for adverse events in left ventricular assist device application. Activation of coagulation showed a time-dependent increase, with no differences between both groups. CONCLUSIONS: This experimental study confirms the hypothesis that axial-flow pumps may induce stronger hemolysis compared to centrifugal pumps, coming along with larger amounts of circulating extracellular vesicles and a stronger PLT activation.

Hemolytic Performance in Two Generations of the Sputnik Left Ventricular Assist Device: A Combined Numerical and Experimental Study.

BACKGROUND: Currently, left ventricular assist devices (LVADs) are a successful surgical treatment for patients with end-stage heart failure on the waiting list or with contraindicated heart transplantation. In Russia, Sputnik 1 LVAD was also successfully introduced into clinical practice as a bridge-to-transplant and a destination therapy device. Development of Sputnik 2 LVAD was aimed at miniaturization to reduce invasiveness, optimize hemocompatibility, and improve versatility for patients of various sizes. METHODS: We compared hemolysis level in flow path of the Sputnik LVADs and investigated design aspects influencing other types of blood damage, using predictions of computational fluid dynamics (CFD) and experimental assessment. The investigated operating point was a flow rate of 5 L/min and a pressure head of 100 mm Hg at an impeller rotational speed of 9100 min-1. RESULTS: Mean hemolysis indices predicted with CFD were 0.0090% in the Sputnik 1 and 0.0023% in the Sputnik 2. Averaged values of normalized index of hemolysis obtained experimentally for the Sputnik 1 and the Sputnik 2 were 0.011 ± 0.003 g/100 L and 0.004 ± 0.002 g/100 L, respectively. CONCLUSIONS: Obtained results indicate obvious improvements in hemocompatibility and sufficiently satisfy the determined miniaturization aim for the Sputnik 2 LVAD development.

Dual-Modality Volume Measurement Integrated on a Ventricular Assist Device.

OBJECTIVE: Ventricular assist devices (VADs) are implanted in patients suffering from end-stage heart failure to sustain the blood circulation. Real-time volume measurement could be a valuable tool to monitor patients and enable physiological control strategies to provide individualized therapy. However, volume measurement using one sensor modality requires re-calibration in the critical time post VAD implantation. METHODS: To overcome this limitation, we have integrated ultrasound and impedance volume measurement techniques into a cannula of an apical VAD. We tested both modalities across a volume range from 140-420 mL using two differently sized and shaped biventricular silicon heart phantoms, which were subjected to physiological pressures in an in-vitro test bench. We compared results from standard calibrated measurements with calculations found by a quadratic optimization for the single modality and their combination (dual-modality) and validated the results using twofold cross-validation. RESULTS: The dual-modality approach resulted in most favorable limits of agreement (LOA) of -0.83 ± 1.54% compared to -13.88 ± 5.90% for ultrasound and -43.45 ± 10.28% for electric impedance, separately. CONCLUSION: The results of the dual-modality approach were as accurate as the standard calibrated measurement and valid over a large range of volumes (140-420 mL). In this in-vitro study, we show how a dual-modality ventricular volume measurement of ultrasound and electric impedance increases the robustness and renders calibration obsolete. SIGNIFICANCE: Ventricular volumes could be measured accurately in the critical period post VAD implantation despite ventricular remodeling.

Dynamic lung behavior under high G acceleration monitored with electrical impedance tomography.

Objective. During launch and atmospheric re-entry in suborbital space flights, astronauts are exposed to high G-acceleration. These acceleration levels influence gas exchange inside the lung and can potentially lead to hypoxaemia. The distribution of air inside the lung can be monitored by electrical impedance tomography. This imaging technique might reveal how high gravitational forces affect the dynamic behavior of ventilation and impair gas exchange resulting in hypoxaemia.Approach. We performed a trial in a long-arm centrifuge with ten participants lying supine while being exposed to +2, +4 and +6 Gx(chest-to-back acceleration) to study the magnitude of accelerations experienced during suborbital spaceflight.Main results. First, the tomographic images revealed that the dorsal region of the lung emptied faster than the ventral region. Second, the ventilated area shifted from dorsal to ventral. Consequently, alveolar pressure in the dorsal area reached the pressure of the upper airways before the ventral area emptied completely. Finally, the upper airways collapsed and the end-expiratory volume increased. This resulted in ventral gas trapping with restricted gas exchange.Significance. At +4 Gx, changes in ventilation distribution varied considerably between subjects, potentially due to variation in individual physical conditions. However, at +6 Gxall participants were affected similarly and the influence of high gravitational conditions was pronounced.

A model-based source separation algorithm for lung perfusion imaging using electrical impedance tomography.

Objective. Electrical impedance tomography (EIT) for lung perfusion imaging is attracting considerable interest in intensive care, as it might open up entirely new ways to adjust ventilation therapy. A promising technique is bolus injection of a conductive indicator to the central venous catheter, which yields the indicator-based signal (IBS). Lung perfusion images are then typically obtained from the IBS using the maximum slope technique. However, the low spatial resolution of EIT results in a partial volume effect (PVE), which requires further processing to avoid regional bias.Approach. In this work, we repose the extraction of lung perfusion images from the IBS as a source separation problem to account for the PVE. We then propose a model-based algorithm, called gamma decomposition (GD), to derive an efficient solution. The GD algorithm uses a signal model to transform the IBS into a parameter space where the source signals of heart and lung are separable by clustering in space and time. Subsequently, it reconstructs lung model signals from which lung perfusion images are unambiguously extracted.Main results. We evaluate the GD algorithm on EIT data of a prospective animal trial with eight pigs. The results show that it enables lung perfusion imaging using EIT at different stages of regional impairment. Furthermore, parameters of the source signals seem to represent physiological properties of the cardio-pulmonary system.Significance. This work represents an important advance in IBS processing that will likely reduce bias of EIT perfusion images and thus eventually enable imaging of regional ventilation/perfusion (V/Q) ratio.

Multi-channel bioimpedance spectroscopy based on orthogonal baseband shifting.

Objective. Multi-channel bioimpedance spectroscopy (BIS) systems typically sample each channel's impedance sequentially using multiplexers and a single analog-to-digital converter. These systems may lose their real-time capability with an increasing number of channels, especially for low excitation frequencies. We propose a new method, called orthogonal baseband shifting (OBS), for high-speed parallel BIS data acquisition at multiple excitation frequencies with low hardware and computational effort.Approach. Similar to orthogonal frequency-division multiplexing, used for digital data transmission, OBS systems use channel-specific orthogonal carrier frequencies to modulate the voltage response of the tissue. Given a suitable choice of carrier frequencies, the modulated signals of all channels sum up without loss of information and cross-talk. The fast Fourier transform (FFT) of the summed signal reveals a spectrum of non-overlapping, interleaved BIS data from which the corresponding BIS data of each channel can be calculated.Main results. In simulations, the system design requires a minimum signal-to-noise ratio of 30 dB to achieve amplitude errors below 1% and phase errors below 0.8°. The hardware realization, called 'AixBIS', has been evaluated for impedance measurements between 0.1 and 10 Ω with multi-frequency excitations between 45 and 180 kHz. The impedance values acquired had an averaged precision of 3.67 mΩ, which is only 0.65‰ in relation to the measured impedance. The phase had a mean precision of 0.46°. Moreover,in vitromeasurements achieved 140 full spectrum acquisitions per second. The impedance change measured in a silicone heart phantom showed a high correlation of 0.83 with the ventricles volume change (flow).Significance. The proposed method enables very fast impedance acquisition of all channels. A complete measurement is performed in the time of a single FFT acquisition, which is equal to the resolution bandwidth of the FFT. In addition, portable and low-power multi-channel BIS devices profit from highly reduced hardware effort. The outstanding performance of OBS measurements with the AixBIS system have the potential forin vivoBIS measurements in real-time.

Monitoring transcellular fluid shifts during episodes of intradialytic hypotension using bioimpedance spectroscopy.

BACKGROUND: Transcellular fluid shifts during dialysis treatment could be related to the frequency and severity of intradialytic hypotension (IDH). We investigated that (i) in addition to ultrafiltration, extracellular fluid (ECF) is further depleted by transcellular fluid shifts and (ii) changes in intracellular fluid (ICF), which have been overlooked so far, or if they were considered, are not understood, might be due to these fluid shifts. METHODS: Thirty-six patients were categorized as haemodynamically stable, asymptomatic IDH or unstable (symptomatic IDH) according to their changes in systolic blood pressure and associated clinical symptoms. Their intradialytic changes in body fluids were studied using bioimpedance spectroscopy measurements and compared among groups. RESULTS: For IDH-prone patients, data showed a rapid drop in ECF that was more than expected from the ultrafiltration rate (UFR) profile and was associated with a significant increase in ICF (P = 0.001). Study of accumulative loss profiles of ECF revealed a loss in ECF up to 300 ml, more than that predicted from UFR for unstable patients. CONCLUSIONS: The considerable discrepancy between the expected and measured loss in ECF might provide evidence of transcellular fluid shifts possibly induced by changes in plasma osmolarity due to haemodialysis. Moreover, the results suggest a pattern of fluid removal in IDH-prone patients that significantly differs from that in haemodynamically stable patients.

Improved Estimation of Left Ventricular Volume from Electric Field Modeling.

Volume measurement is beneficial in left ventricular assist device (LVAD) therapy to quantify patient demand. In principle, an LVAD could provide a platform that allows bioimpedance measurements inside the ventricle without requiring additional implants. Conductance measured by the LVAD can then be used to estimate the ventricular radius, which can be applied to calculate ventricular volume. However, established methods that estimate radius from conductance require elaborate individual calibration or show low accuracy. This study presents two analytical calculation methods to estimate left ventricular radius from conductance using electric field theory. These methods build on the established method of Wei, now considering the dielectric properties of muscle and background tissue, the refraction of the electric field at the blood-muscle boundary, and the changes of the electric field caused by the measurements. The methods are validated in five glass containers of different radius. Additional bioimpedance measurements are performed in in-vitro models that replicate the left ventricle's shape and conductive properties. The proposed analytical calculation methods estimate the radii of the containers and the in-vitro models with higher accuracy and precision than Wei's method. The lead method performs excellently in glass cylinders over a wide range of radii (bias: 1.66%-2.48%, limits of agreement < 16.33%) without calibration to specific geometries.

Bandwidth and Common Mode Optimization for Current and Voltage Sources in Bioimpedance Spectroscopy.

Bioimpedance measurements use current or voltage sources to inject an excitation signal into the body. These sources require a high bandwidth, typically from 1 kHz to 1 MHz. Besides a low common mode, current limitation is necessary for patient safety. In this paper, we compare a symmetric enhanced Howland current source (EHCS) and a symmetric voltage source (VS) based on a non-inverting amplifier between 1 kHz and 1 MHz. A common mode reduction circuit has been implemented in both sources. The bandwidth of each source was optimized in simulations and achieved a stable output impedance over the whole frequency range. In laboratory measurements, the output impedance of the EHCS had its -3 dB point at 400 kHz. In contrast, the VS reached the +3 dB point at 600 kHz. On average over the observed frequency range, the active common mode compensation achieved a common mode rejection of -57.7 dB and -71.8 dB for the EHCS and VS, respectively. Our modifications to classical EHCS and VS circuits achieved a low common mode signal between 1 kHz and 1 MHz without the addition of complex circuitry, like general impedance converters. As a conclusion we found VSs to be superior to EHCSs for bioimpedance spectroscopy due to the higher bandwidth performance. However, this only applies if the injected current of the VS can be measured.

Car Seats with Capacitive ECG Electrodes Can Detect Cardiac Pacemaker Spikes.

The capacitive electrocardiograph (cECG) has been tested for several measurement scenarios, including hospital beds, car seats and chairs since it was first proposed. The inferior signal quality of the cECG compared to the gold standard ECG guides the ongoing research in the direction of out-of-hospital applications, where unobtrusiveness is sought and high-level diagnostic signal quality is not essential. This study aims to expand the application range of cECG not in terms of the measurement scenario but in the profile of the subjects by including subjects with implanted cardiac pacemakers. Within this study, 20 patients with cardiac pacemakers were recruited during their clinical device follow-up and cECG measurements were conducted using a seat equipped with integrated cECG electrodes. The multichannel cECG recordings of active unipolar and bipolar pacemaker stimulation were analyzed offline and evaluated in terms of Fß scores using a pacemaker spike detection algorithm. Fß scores from 3652 pacing events, varying from 0.62 to 0.78, are presented with influencing parameters in the algorithm and the comparison of cECG channels. By tuning the parameters of the algorithm, different ranges of Fß scores were found as 0.32 to 0.49 and 0.78 to 0.88 for bipolar and unipolar stimulations, respectively. For the first time, this study shows the feasibility of a cECG system allowing health monitoring in daily use on subjects wearing cardiac pacemakers.

In Silico and in Vitro Conductivity Models of the Left Heart Ventricle.

Ventricular Assist Devices (VADs) are used to treat patients with cardiogenic shock. As the heart is unable to supply the organs with sufficient oxygenated blood and nutrients, a VAD maintains the circulation to keep the patient alive. The observation of the patient's hemodynamics is crucial for an individual treatment; therefore, sensors to measure quantifiable hemodynmaic parameters are desirable. In addition to pressure measurement, the volume of the left ventricle and the progress of muscle recovery seem to be promising parameters. Ongoing research aims to estimate ventricular volume and changes in electrical properties of cardiac muscle tissue by applying bioimpedance measurement. In the case where ventricular insufficiency is treated by a catheter-based VAD, this very catheter could be used to conduct bioimpedance measurement inside the assisted heart. However, the simultaneous measurement of bioimpedance and VAD support has not yet been realized, although this would allow the determination of various loading conditions of the ventricle. For this purpose, it is necessary to develop models to validate and quantify bioimpedance measurement during VAD support. In this study, we present an in silico and an in vitro conductivity model of a left ventricle to study the application of bioimpedance measurement in the context of VAD therapy. The in vitro model is developed from casting two anatomical silicone phantoms: One phantom of pure silicone, and one phantom enriched with carbon, to obtain a conductive behavior close to the properties of heart muscle tissue. Additionally, a measurement device to record the impedance inside the ventricle is presented. Equivalent to the in vitro model, the in silico model was designed. This finite element model offers changes in material properties for myocardium and the blood cavity. The measurements in the in vitro models show a strong correlation with the results of the simulation of the in silico model. The measurements and the simulation demonstrate a decrease in impedance, when conductive muscle properties are applied and higher impedances correspond to smaller ventricle cross sections. The in silico and in vitro models are used to further investigate the application of bioimpedance measurement inside the left heart ventricle during VAD support. We are confident that the models presented will allow for future evaluation of hemodynamic monitoring during VAD therapy at an early stage of research and development.