Jacobs Journal of Anesthesiology and Research

Novel Method for Continuous Long-Term Monitoring of Animals Via Multispectral Photoplethysmography

*Julia T. Rettberg
Department Of Aquatic Animal Medicine, TUM School Of Medicine, Munich, Germany

*Corresponding Author:
Julia T. Rettberg
Department Of Aquatic Animal Medicine, TUM School Of Medicine, Munich, Germany
Email:j.rettberg@swiss-kiss-petrol.ch

Published on: 2018-11-19

Abstract

Objectives: To investigate the reliability of vital sign parameters in different animal species measured with a Biovotion vital sign monitor (vsm-1; Biovotion AG, Switzerland) and to introduce a novel method for continuous long-term monitoring of freely moving animals using different medical indications. Study Design: All measurements were performed in animals anesthetized in the context of different preclinical studies. The interventions included anesthesia and intubation exercises, endoscopies, abdominal surgeries, ophthalmic surgeries, and cardiac studies. Animals: Eleven domestic pigs, ten minipigs and sixteen sheep. Methods: Heart rate (HR) and blood oxygenation (SpO2) were simultaneously measured under general anesthesia with conventional monitoring systems 1-3 using the vsm-1 at eight different positions. Data from both parameters were then compared statistically. Real-time values of various additional vital signs, including skin temperature, skin blood perfusion index, blood pulse wave, and motion/ acceleration, were transmitted from the vsm-1 to a laptop computer via Bluetooth and interpreted with the appropriate engineering tool (pyWatch; Biovotion AG, Switzerland). Results: The HR values measured using the vsm-1 (HRVSM) did not significantly differ from those measured simultaneously using conventional monitoring systems 1-3 (HRConventional). No relevant systematic differences (Δ HR) regarding the animal species or the position of the vsm-1 and no association with the magnitude of the measurements (Mean HR) were detected. The distribution of the differences showed no notable trends. The extent of the differences in the means (Δ Mean HR) related to individual positions was not consistent among the species. SpO2 measured using the vsm-1 provided reliable mean values in only one-third of the measurement cycles. Conclusions and Clinical Relevance: Parameter visualization via a user-friendly, real-time interface provides important information for potential applications using the vsm-1 for long-term monitoring. Further development of the investigated parameters, particularly SpO2, and others will improve both the monitoring and assessment of an animal’s well-being.

Keywords

Heart Rate; Monitoring; Cardiovascular; Photophletysmography; Infrared; Microcirculation; Intensive Care; Pain Assessment; Stress; Animal Welfare

Introduction

The primary diagnostic challenge in veterinary medicine compared with human medicine is the fact that animals cannot speak. Adequate knowledge about the patient depends on the anamnesis given by its owner and on a thorough clinical examination. Furthermore, stress, anxiety and aggressiveness not only make the examination more difficult but also disguise symptoms and may induce false investigative results. This observation was exemplified by Bragg et al, who compared measurements of panting and different vital sign parameters in healthy dogs in a veterinary clinic with measurements taken in a familiar environment.

Moreover, farm animals are uncomfortable when separated from their herd, flock or housing group. Kuwahara et al. [2] investigated the impact of altered husbandry conditions of minipigs on their stress response and concluded that the animals returned to normal values only after 14 days. Above all, an assistant forcing an animal to keep still for a common examination goes against the nature of the animal and causes stress. Reference values for heart rates (HR) listed in textbooks [3] can reflect impossible levels of quiescence.Biovotion AG of Zurich, Switzerland, developed a device for human medicine that provides reliable vital sign data for monitoring patients at home and in hospitals (Figure 1). The Biovotion vital sign monitor (vsm-1) implements multispectral photoplethysmography (PPG) for vital sign measurements. This technique uses the interaction of transmitted green, red and infrared light with the dermal tissue and vessels to estimate changes in the blood volume [4]. The penetration depth, extent of scattering and absorption depend on the wavelength of the light and the amount of the reflected light detected by the photodetector [5]. Current parameters monitored by the vsm-1 include HR, blood oxygenation (SpO2 ), skin temperature (skinT), skin blood perfusion index (PI), blood pulse wave (BPw) and motion/acceleration (MA). Additional parameters are expected to follow in the future [6]. Further processing of the collected data allows the identification of stress and sleep disorders in humans [7].

Figure 1. The Biovotion vital sign monitor (vsm-1). The sensors of the vsm-1 include an optical sensor (1), an acceleration sensor, a temperature sensor and a sweat sensor (2).

Previous attempts to establish alternative methods for HR monitoring in animals have resulted in mixed outcomes. Marchant-Forde et al. [8] used a Polar RR Recorder in pigs but had to address artifacts caused by movement and muscle activity. Similar studies in dogs concluded that a related HR monitor could replace electrocardiography (ECG) when the activity level of the animals would not disturb the measurement [9]. However, the saved data had to be carefully and critically reviewed by the examiner [10]. With the Holter ECG as the gold standard method for long-term cardiac monitoring, several studies have tested these devices in different animals [11-13]. The practical application of these devices, including the use of a bandage to prevent the equipment from being damaged, has been described as bothersome compared to the custom-made jackets used in earlier studies [11], although such jackets were recommended only for a certain number of animals for financial reasons [12]. Nevertheless, the medical equipment marketplace does not yet offer any device for monitoring vital sign parameters other than the HR as a matter of routine in animals. The purpose of this study was to investigate the level of agreement between HR and SpO2 values recorded in animals using the vsm-1 under general anesthesia compared with conventional monitoring systems.1-3 Additionally, the assessment of the reliability of the additional parameters currently being measured by the vsm-1 is focused on the possibility of introducing a novel method for continuous long-term monitoring of freely moving animals using different medical indications.

Materials and Methods

Animals

The study included eleven domestic pigs (8 boars, 6 of them castrated, and 3 sows) with weights of 55.2±26.8 kg, ten castrated minipig boars with weights of 50.8±16.2 kg, and fourteen adult female sheep with weights of 76.25±8.25 kg.

Study Design

All measurements were approved by the local government and were performed in accordance with the 2010/63/EU directive as well as the transacted German law (TierSchG, TierSchVersV). The animals were anesthetized in the context of different preclinical studies in the Center for Preclinical Research at the TUM School of Medicine, Munich, Germany. For domestic pigs, the interventions included anesthesia and intubation exercises, laparoscopies, and abdominal surgeries. All measurements in the minipigs were performed over the course of an ophthalmic surgery, whereas the sheep were part of an extensive cardiac study. 

Preparations for Measurement

Each pig was premedicated with a mixture of azaperone, ketamine, and atropine sulfate (2.0+0.02+15.0 mg/kg IM) to prevent any profuse salivation from hindering intubation [14,15]. Anesthesia was induced by intravenous bolus injection of 1% propofol as required for adequate intubation (~1.5 mg/ kg). Given the risk of malignant hypothermia [15], the pigs were ventilated only with ambient air and oxygen, and no inhalation anesthetic was used. Therefore, anesthesia was maintained by intravenous drip of 2% propofol (12.5 mg/ kg/h) according to the guidelines provided in specialist literature [16]. An alternative form of general anesthesia with a pentobarbital sodium (2.5 mg/kg/h) drip was chosen for the surgeries in the final studies. The compiled vital sign data did not differ significantly between the two different forms of anesthesia. Premedication of the sheep was achieved by a mixture of xylazine and atropine sulfate (0.4+0.0065 mg/kg IM) for a stable HR [15], and the induction of general anesthesia was administered by intravenous bolus of 2% propofol as required for an adequately feasible intubation (~3.0-5.0 mg/ kg). To maintain general anesthesia, the sheep were ventilated with 1.5% isoflurane. Operative pain management using fentanyl (0.015 mg/kg IV in intervals of 2030 min) was used equally for all animal species. Additional medications included metamizole sodium (40.0 mg/kg IV), ketamine (8.0 mg/kg/h IV), and an intercostal nerve block (2.0 ml of a mixture of 2%-lidocaine hydrochloride and 0.75%-bupivacaine for each intercostal plane between the 3rd-5th rib at a maximum dose of 20.0 ml per animal) for the sheep cardiac surgeries.

POSITIONS OF THE DEVICE

After shaving and cleaning the appropriate positions (Figure 2) for the vsm-1, the device was fitted by a modified.

Figure 2. Positions of the vsm-1. The vsm-1 was attached either by custom-made hook-and-loop fastener belts or by a strip of adhesive tape in the following positions: thoracic spine caudal to the scapulae (1), dorsal cervical spine (2), external jugular vein (3), atop the sternum near the cardiac apex (4), caudal of the elbow joint (5), dorsal lumbar spine (6), abdomen (7), and the inner thigh (8).

DATA ACQUISITION

The optical sensor of the vsm-1 (Figure 1), consisting of 3 LEDs emitting light in the green, red and infrared wavelengths and a photodetector detecting the light reflected from the tissue, provided real-time monitoring of HR, SpO2, PI, and BPw. Additional sensors, including an acceleration sensor, a temperature sensor, and a sweat sensor, enabled the measurement of skinT and MA (in this context, any movement of the device or the animal wearing it). The data were transferred from the vsm-1 to a laptop computer using a Bluetooth connection and moniversion of the enclosed fastening system and a range of custom-made hook-and-loop fastener belts of various lengths. When the abdomen and inner thigh positions were used, the vsm-1 was attached using a strip of adhesive tape.tored in real-time with the appropriate engineering tool (pyWatch; Biovotion AG, Switzerland). These values were frequently noted in an extended anesthesia chart together with the vital parameters, which were written down at the same time from the displays of the conventional monitoring systems.1-3 Additionally, the vsm-1 data were stored as text files and transferred to the computer for further processing. Vital sign values were excluded from the protocol for at least 5 min subsequent to major manipulations, such as bedding the animal in an appropriate position or the administration of drugs with the potential to cause hemodynamic changes [17].

MEASUREMENT CYCLES

Each measurement cycle was composed of the simultaneously measured HR noted at 10-s intervals from the pyWatch (HRVSM) and the conventional monitoring systems 1-3 (HRConventional). The cycle was repeated three or more times, each using different positions. Within a measurement cycle, the mean was calculated both from HRVSM and from HRConventional. The number of the mean HRs for each animal is the sum of all measurement cycles per animal and positions. As a result of the minor agreement between the SpO2 values of the compared measurement methods, the data were interpreted differently. If the mean of the measured SpO2 values using the vsm-1 within a measurement cycle was 95% or higher, the cycle was considered reliable.

STATISTICAL ANALYSIS

The statistical method proposed by Martin Bland and Altman [18] for assessing the agreement of measurement methods was implemented to compare the HRVSM and the HRConventional. Differences in the mean HRs (Δ HR [VSM, Conventional]) were plotted against their mean values (Mean HR [VSM, Conventional]) to determine the presence of systematic differences, the distribution of the differences, and an association of the magnitude of the measurements with the differences (Figure 3). Symbols and colors differentiate species and indicate whether reliable values of SpO2 ≥95% were measured during each measurement cycle. Additional lines represent the standard deviation (STD) and the upper and lower limits of agreement (mean±1.96×STD), which were calculated comprehensively using the values for all animals.

Figure 3. Scatter plot of the differences in the means of the heart rates measured using the two methods (Δ HR [VSM, Conventional]) plotted against the corresponding means (Mean HR [VSM, Conventional]). Symbols and colors differentiate species and indicate whether reliable values of blood oxygenation (≥95%) were measured during each measurement cycle. Additional lines represent the standard deviation (STD) and the upper and lower limits of agreement (mean±1.96×STD), which were calculated using the combined values for all animals. Differences in the means (Δ Mean HR) and their STD were compared considering all measurements in individual animal species (Table 1), all positions in all animal species (Table 2), or different positions within each animal species (Table 3).

THE SPIRAL VIEW IN THE APPLICATION

Secondary to the option of real-time monitoring by numerical values, pyWatch visualizes the combination of all measured vital sign parameters in a special spiral view (Figure 4) documenting the time span of the measurement period. The minute is displayed in a manner analogous to a clock face in the outer circle, starting when the device registers skin contact with the PPG sensor. While the minute interface is overwritten every hour, the hour is filled in on the prescribed inner spiral form. The color of the inner bar is associated with defined HR intensities. Turquoise indicates low HR (110 bpm). Any position change of the device is registered as MA and visualized as an extension of the inner bar. Even very remote shifting of the animal during surgery produces slight extensions. The color of the outer bar is associated with the BPw. The numerical values of this parameter, which does not have a unit, range from 0-5 and are visualized in shades from light to dark gray/blue or black. This range provides the examiner with an estimate of the level of relaxation or stress [7].

Figure 4. Spiral view in the application. The combination of all measured vital sign parameters using the vsm-1 during the time span of a measurement period are visualized in the spiral. The minute is displayed in a manner analogous to a clock face in the outer circle, starting when the device registers skin contact through the PPG sensor. While the minute interface is overwritten every hour, the hour fills the prescribed inner spiral form. The color of the inner bar is associated with defined intensities of the heart rate (HR): turquoise (< 60 bpm), green (60-70 bpm), yellow (~80 bpm), orange (90 100 bpm), red (~100 bpm), and very dark red (>110 bpm). Any change in the device position is registered as motion/ acceleration (MA) and visualized as extensions of the inner bar. Even very remote shifting of the animal during surgery produces slight extensions. The color of the outer bar is associated with the blood pulse wave (BPw). The numerical values of this parameter, which does not have a unit, range from 0-5 and are visualized in shades from light to dark gray/blue or black. This range provides the examiner with an estimate of the level of relaxation or stress [1].

Symbols:

A: * intravenous injections of fentanyl

† incision in the animal’s skin

B:* position changes (extensions of the inner bar)

† the animal is turned from a lateral to a dorsal position

§ blood pressure, measured by invasive means

C:* boluses of fentanyl every 20 min (inner spiral)

† last two injections of fentanyl (minute interface)

D: * boluses of fentanyl (0.015 mg/kg)

† restart of the vsm-1

‡ drip of dopamine (maximum HR 178 bpm)

§ stop of the drip

? euthanasia (pentobarbital 50.0 mg/kg IV and potassium chloride)

Results:

Δ HR [VSM, Conventional] ranged from -1.42 to 1.83 bpm. We did not observe any relevant systematic differences regarding animal species or positions, although the HRVSM was slightly higher than the HRConventional (STD 0.42 bpm). Limits of agreement settled at 0.7 bpm and 1.54 bpm. Outliers in the scatter plot could not be assigned to specific HR intensities, suggesting a lack of an association with the magnitude of the mean values. Moreover, the distribution of the Δ Mean HR considering particular animals of one species did not appear to differ from the distribution in the other animals (Figure 3). The extent of the Δ Mean HR related to individual positions was not consistent among the species. Using the caudal elbow as an example, the difference was 0.30 when calculated comprehensively (Table 2), but it was 0.42 when only the data from domestic pigs were evaluated, 0.24 for minipigs, 0.28 for sheep (Table 3). In contrast, the position atop the sternum had a comprehensive Δ of 0.48, but it was 0.65 in minipigs, and 0.72 in sheep. STDs ranged from 0.31 - 0.68 and didi not correlate negatively or positively with the appropriate Δ Mean HR. Mean SpO2 values where ≥95% in 34.3% of all measurement cycles (Figure 3Figure 3). Differentiating the animal species and positions, measurement cycles in minipigs were considered reliable in 48.7% of the cycles and 36.3% of the cycles in domestic pigs, whereas sheep exhibited SpO2≥95% in 17.5% of the cycles.

INTERPRETATION OF THE SPIRAL – INNER BAR

The spirals drawn from the collected vsm-1 data exhibited colors from turquoise to very dark red (Figure 4). No tendencies for certain color shades were noted in particular species. The minipigs appeared to exhibit similar colors. The majority of the animals had an HR of 60-90 bpm (green to yellow). High color diversity was obtained in sheep. One had HRs <60 bpm, whereas the others varied between 60-175 bpm  within a similar length of time. Individual sheep had higher HRs at the start of measurement immediately after the induction of anesthesia, producing spirals with a color spectrum from orange to dark red and changing back to yellow whenever analgesics were administered (Figure 4A*). In addition to the large extensions and narrow gaps, which indicate position changes (Figure 4B*), the shifting of the animal from a lateral to dorsal position (†) could be visualized as an abrupt change in color from yellow to green. The simultaneously measured mean arterial pressure (Figure 4B§) changed analogously (69.3/61.5 mmHg). Rather slight extensions of the inner bar were produced in the context of surgeries with continuous manipulations. Figure 4D exemplifies the effect of drug-induced tachycardia up to 187 bpm.

INTERPRETATION OF THE SPIRAL – OUTER BAR

The colors of the inner and outer bars did not correlate. Low HR (turquoise) was combined with dark outer bars, and average HR (yellow) could have light or dark outer bars. BPw was not correlated with the length of measurement. The outer bar initially was very dark in the spirals of sheep with high HRs at the start of measurement and became a lighter gray/blue over half an hour. In contrast, the outer bar changed to dark gray/ blue over time in the spiral of another specimen (Figure 4).

LINE GRAPHS OF THE VSM-1 DATA

Saved vsm-1 data can also be analyzed in common line graphs. The data from Figure 4, B+D are presented in Figures 5 & 6.

Figure 5. Interpretation of the numerical vsm-1 data from spiral B. Extensions of MA due to position changes (*) and gaps due to loss of skin contact are visible in the graph. The altered circulatory function after turning the animal from a lateral to a dorsal position (†), described by the spiral view, is likewise presented by the HR values. After a bolus of fentanyl (‡), HR decreases more. The invasively measured systolic (SAP), diastolic (DAP) and mean arterial pressure (MAP) were documented continuously and marked three times (§) for exemplification before and after the shift of the animal into the dorsal position. Values changed from Ø 91.8 mmHg, SAP/59.0 mmHg, and DAP/69.3 mmHg MAP to Ø 80.4 mmHg SAP, 51.1 mmHg DAP, and 61.5 mmHg MAP after 5 min and to Ø 90.4 mmHg SAP, 68.4 mmHg DAP, and 76.3 mmHg MAP 15 min after turning the animal.

The simultaneously measured HRConventional (not shown in the graph) exhibited a Δ-0.2 bpm from the HRVSM over the analyzed time span of 75 min. Additionally, the heart rate quality (HRQ) is indicated by the thickness of the HR line in the line graph. HRQ>50% is presented as a bold HR line. SpO2 was ≥95% in 49.2% of the measurements (Figure 5A). Moreover, the quality of SpO2 measurement was analyzed (SpO2Q), and data are presented as a bold line if ≥50%. The bold line comprises 35.4% of the SpO2 values. PI increased shortly after each position change* and fitting of the vsm-1 and slowly decreased during the subsequent measurement period. The skinT appears to initially increase rapidly, followed by a slower approach to values of ~36°C. Neither a correlation with other measured vital sign parameters nor that with rectal temperature was observed, and the temperature was taken at periodic intervals (not included in the graph). BPw ranged from 0-2.5 (multiplied with 100 without unit, Figure 5B) and initially exhibited values of ~1.0 in the first 53 min (see also Figure 4B). BPw quickly increased to greater amplitudes (150- 250 in the graph) during the last 20 min. MA maxima, also unitless, settled at ~100 (Figure 5C) in six position changes and exhibited slight extensions between them. The additionally measured parameter, BARO, was included in the graph but not further analyzed. The HRQ stayed in a range of 60-100% and fell to <50% only when the device lost skin contact*.

Figure 6. Interpretation of the numerical vsm-1 data from spiral D.

Discussion:

Statistical comparison of HRVSM and HRConventional agreed with the subjective impression at the time of monitoring. There were small variations in consecutive HRs, and Δ HR [VSM, Conventional] was unremarkable. If the HR intensity changed over time, both devices registered the drift simultaneously and transmitted either higher or lower values. During such a visually registered shift, variations were initially greater and the Δ HR was more prominent for some minutes until the HR settled either at the higher or lower intensity. The impossibility of any unambiguous assignment of outliers in the scatter plot suggests that there was no position or animal species that fulfilled the requirements for HR measurement using the vsm-1 better than another. Given that 40% of the mean HRs were in the range of 75-85 bpm, the distribution of the Δ HR appears to be slightly more apparent in the middle of the scatter plot.

However, higher mean HRs of 90-100 bpm also exhibited a margin of deviation between 1.0 and 1.36 bpm. This finding confirms that the differences occurred independently of the magnitude of the measurements. Low HRs of 60-70 bpm were measured with a reliability equal to that of higher or very high HRs. In fact, even the maximum deviation observed in this study was not interpreted as clinically relevant. Regarding the fitting of the device, domestic pigs were easy to handle because of their human-like skin, which required the least effort to prepare (shaving and cleaning), whereas the hard bristles of the minipigs, in addition to their often dirty and slightly pigmented rigid skin, were much more bothersome. In contrast, the oily skin of the sheep and the constitution of their connective tissue made fitting the equipment subjectively more difficult. Nevertheless, differences in means and STDs of the different positions in sheep were not exceedingly different from those in other species, as described above. The reason for the poor reliability of the SpO2 measurement by vsm-1 could not be found. The initial theory, that the signals of the sensors could be interfering, could not be confirmed. SpO2Q did not correlate positively or negatively with the HRQ during a measurement period. This parameter requires further development before using the vsm-1 with medical indications requiring SpO2 monitoring.

INTERPRETATION OF THE SPIRAL – INNER BAR

Differences in color diversity among species could be explained by the similarity of the surgeries in the minipigs and the higher pain level associated with the cardiac surgeries of the sheep. Additionally, the sheep had a far higher agitation level in advance of the induction of anesthesia. The appearance of medical personnel in the housing area of the animals was met with curiosity in domestic pigs and minipigs, whereas sheep reacted with signs of stress when they had to be separated from the herd for the application of sedatives and induction of anesthesia. The precise visualization of MA in real-time, in this context caused by passive movement of the device or the animal wearing it, also suggests the possibility of future monitoring of freely moving animals. The extent to which measurement might be hindered by motion/ skin movement during maximal pace requires further investigation. Ultimately, the exhibition of the altered cardiac activity after shifting the animal (Figure 4B) facilitated the monitoring of anesthetized animals. The explanation of the abrupt change in HR after bringing the animal to dorsal recumbency is found in the circulatory assimilation, which is needed because the full body weight was then resting on the cava veins. This fact was also confirmed by the documented mean arterial pressure (MAP)before/after the position change. HR changes in connection with the application of analgesics were interpreted as a counteraction of pain as a stimulus for an increased HR with the drug-induced pain relief. This might be useful as a guide for veterinarians to identify the most adequate dosage for sufficient pain management in conjunction with the future parameter, heart rate variability (HRV), which we will soon be able to measure [6].

INTERPRETATION OF THE SPIRAL – OUTER BAR

 According to Biovotion engineers, BPw is associated with HRV. Therefore, when the time intervals between consecutive heart beats are very regular and instantaneous HRs are quite constant, BPw provides higher values, as indicated by the dark blue/black color of the outer bar in the spiral (Figure 4). In general, a higher stimulation of the sympathetic nervous system during the induction of anesthesia and before the application of analgesics might be an explanation for the dark color of the outer bar (Figure 4A). This result is in accordance with investigations of stress levels using vsm-1 in human beings [7]. Based on the color change to a lighter shade of gray/blue after complete induction of general anesthesia (with application of analgesics), BPw values probably slowly decreased in a manner negatively correlation with a higher HRV. As noted in the introduction, the vsm-1 is still in development, and the interpretation of the BPw with the help of the HRV will be enhanced in the near future. A recent study of rats investigated the consequences of stress with different approaches. In that study, Park et al. [19] demonstrated a weakened ability of the organism to.

PERSPECTIVES FOR VETERINARY MEDICINE

After further development of the device and adjustment of the algorithm to the requirements of veterinary medicine, the indications for long-term monitoring by the vsm1 might facilitate 24-h surveillance of animals in intensive care units or the control of epileptic seizures in dogs. The device might better provide responses to nociception in anesthetized animals or alleviate sleep disorders in brachycephalic dogs [21,22]. Moreover, extensive care for pregnant mares and colic detection would be principal focuses of further studies in large animals. If the connection between changes in the HRV and emotions in pets [23] can be confirmed by measurements using the vsm-1, new fields for behavioral studies will be opened in the future.

Footnotes:

1. Ohmeda Patient Monitor, Datex Ohmeda GmbH, Duisburg, Germany

2. G3 Handheld Pulse Oximeter, Goldway, China

3. Lifepak Multiparameter Monitor, Physio-Control, Inc. Redmond, WA, USA

Acknowledgements:

This study was supported by Biovotion AG in Zurich, Switzerland. The authors thank Dr. Andreas Caduff for the provision of vsm-1a devices and Wolfgang Werner for moderating updates and contact persons. Special thanks to Biovotion engineers Ursin Tuor, Dr. Mattia Zanon, Dr. Thomas Degen and Stephan Bachofen for their extensive technical assistance.

Conflicts of interest:

The authors declare that there are no conflicts of interest.

References:

1. Bragg RF, Bennett JS, Cummings A, Quimby JM. Evaluation of the effects of hospital visit stress on physiologic variables in dogs. J Am Vet Med Assoc. 2015, 246(2): 212-215.

2. Kuwahara M, Tsujino Y, Tsubone H, Kumagai E, Tsutsumi H, Tanigawa M. Effects of pair housing on diurnal rhythms of heart rate and heart rate variability in miniature swine. Exp Anim. 2004, 53(4):303-309.

3. Baumgartner W: Allgemeiner Untersuchungsgang. In Klinische Propädeutik der Haus- und Heimtiere, 7 edition. Edited by Baumgartner W. Stuttgart: Parey; 2009:82-89.

4. Allen J. Photoplethysmography and its application in clinical physiological measurement. Physiol Meas. 2007, 28(3): R1-39.

5. Maeda Y, Sekine M, Tamura T. The advantages of wearable green reflected photoplethysmography. J Med Syst. 2011, 35(5): 829-834.

6. Biovotion. Cinical grade vital sign data. Current parameters. 2017, http://www.biovotion.com/. 2018.

7. Biovotion. Get started. First steps. The Biovotion “Spiral” explained. 2017, https://biovotion.zendesk.com/ hc/en-us/articles/213542089-The-Spiral-View-in-theApp 2018.

8. Marchant-Forde RM, Marlin DJ, Marchant-Forde JN. Validation of a cardiac monitor for measuring heart rate variability in adult female pigs: accuracy, artefacts and editing. Physiol Behav. 2004, 80(4): 449-458.

9. Essner A, Sjostrom R, Ahlgren E, Lindmark B. Validity and reliability of Polar(R) RS800CX heart rate monitor, measuring heart rate in dogs during standing position and at trot on a treadmill. Physiol Behav. 2013, 114- 115:1-5.

10. Essner A, Sjostrom R, Ahlgren E, Gustas P, Edge-Hughes L, Zetterberg L et al. Comparison of Polar(R) RS800CX heart rate monitor and electrocardiogram for measuring inter-beat intervals in healthy dogs. Physiol Behav. 2015, 138: 247-253.

11. Suzuki A, Tsutsumi H, Kusakabe K, Kuwahara M, Sugano S, Tanigawa M. Establishment of a 24-hour electrocardiogram recording system using a Holter recorder for miniature swine. Lab Anim. 1998, 32(2): 165-172.

12. Scheer P, Svoboda P, Sepsi M, Janeckova K, Doubek J. The electrocardiographic Holter monitoring in experimental veterinary practice. Physiol Res. 2010, 59 Suppl 1:S59-64.

13. Rasmussen CE, Vesterholm S, Ludvigsen TP, Haggstrom J, Pedersen HD, Moesgaard SG et al. Holter monitoring in clinically healthy Cavalier King Charles Spaniels, Wire-haired Dachshunds, and Cairn Terriers. J Vet Intern Med. 2011, 25(3): 460-468.

14. Kersjes CM, Mack AC, Böttcher P, Henke J, Erhardt W, Heinritzi K. Anaesthetic methods for pigs suitable in daily practice. Tierarztl Prax Ausg G Grosstiere Nutztiere 2000, 28:225-232.

15. Erhardt W, Henke J, Baumgartner C: Labortiere. In Anästhesie und Analgesie beim Klein-und Heimtier: mit Exoten, Labortieren, Vögeln, Reptilien, Amphibien und Fischen, 2nd edition. Edited by Erhardt W, Henke J, Haberstroh J, Baumgartner C, Tacke S. Stuttgart: Schattauer Verlag; 2012:782-786.

16. Anderson DE, St. Jean G: Anesthesia and surgical procedures in swine. In Diseases of Swine, 10th edition. Edited by Zimmerman JJ, Karriker LA, Ramirez A, Schwartz KJ, Stevenson GW. Chichester: Wiley-Blackwell; 2012:122.

17. Pfeiffer N: Untersuchungen zur akuten kardiovaskulären und hämodynamischen Wirkung der i.v. injizierbaren Hypnotika Propofol und Alfaxalon beim Schwein. Munich, Germany, DVM thesis, Ludwig-Maximilians-Universität München, 2015.

18. Martin Bland J, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. The Lancet. 1986, 327:307-310.

19. Park SE, Park D, Song KI, Seong JK, Chung S, Youn I. Differential heart rate variability and physiological responses associated with accumulated short- and longterm stress in rodents. Physiol Behav. 2017, 171:21-31.

20. Hwang HW, Hyun HK, Kim YJ, Kim JW, Shin TJ. Evaluation of stress response during mesiodens extraction under general anesthesia using heart rate variability. J Oral Maxillofac Surg. 2017, 75(5): 931-937.

21. Roedler FS, Pohl S, Oechtering GU. How does severe brachycephaly affect dog’s lives? Results of a structured preoperative owner questionnaire. Vet J. 2013, 198(3): 606-610.

22. Hendricks JC, Kline LR, Kovalski RJ, O’Brien JA, Morrison AR, Pack AI. The English bulldog: a natural model of sleep-disordered breathing. J Appl Physiol (1985) 1987, 63:1344-1350.

23. Zupan M, Buskas J, Altimiras J, Keeling LJ. Assessing positive emotional states in dogs using heart rate and heart rate variability. Physiol Behav 2016, 155:102-111.

Table 1. Evaluation of the differences (Δ) in the means and their standard deviations considering all measurement cycles comprehensively for individual animal species.

Table 2. Evaluation of differences (Δ) in the means and their standard deviations considering all measurement cycles at each position comprehensively for all animal species.

Table 3. Evaluation of differences (Δ) in the means and their standard deviations considering all measurement cycles at different positions for each animal species.