Abstract
Brachycephalic obstructive airway syndrome (BOAS) is associated with significant morbidity and mortality. Routine clinical evaluation fails to detect physiologic consequences of BOAS including airflow limitation, exercise intolerance, and impaired thermoregulation. A six-minute walk test (6MWT) with infrared thermography (IRT) may aid detection and clinical management by assessing the physiologic consequences of BOAS. IRT has been used in dogs to assess thermoregulation and in people with obstructive sleep apnea. Our objectives were to compare 6MWT and IRT parameters between healthy mesaticephalic (Mesa) and brachycephalic (Brachy) dogs, and dogs with BOAS. 6MWT parameters include normalized distance walked (ND), rectal temperature, pulse, respiratory rate, and pulse oximetry (SPO2). Mean (Tmean) and maximum (Tmax) IRT temperatures at 3 regions of interest (ROI) were evaluated. Evaluation timepoints were pre-6MWT, immediately post-6MWT (T0) and 5 (T5) and 15min post-6MWT (T15). No significant difference in ND, SPO2, or temperature were found between groups (p>.05). BOAS dogs had higher dorsal and rostral Tmax and Tmean temperatures compared to Mesa dogs at all timepoints (p < .05). BOAS dogs had higher Tmean temperatures compared to Brachy dogs at baseline and T15 and T5 and T15 for dorsal and rostral ROIs respectively (p < .001). ROC analysis showed significant discrimination between BOAS and non-BOAS (Brachy and Mesa) dogs with areas under the curve between 0.79–0.96. Significant moderate correlations were found between IRT temperatures, ND and rectal temperature. This pilot study demonstrates the potential in pairing the 6MWT and IRT with evaluation of clinical signs as screening tool to identify dogs with BOAS.
Introduction
Brachycephalic obstructive airway syndrome (BOAS) is a clinically significant and potentially life-threatening condition in dogs [1, 2]. Brachycephalic obstructive airway syndrome occurs because of a spectrum of anatomic abnormalities inclusive of, stenotic nares, hyperplastic soft palate, laryngeal collapse, aberrant conchal growth, macroglossia, and hypoplastic trachea [3, 4]. These anatomic changes result in increased upper airway resistance, reduced and turbulent airflow, and upper airway inflammation [3–5]. These predispose affected dogs to aerodigestive and obstructive sleep disorders, and potentially life-threatening upper airway obstruction [6–9]. Furthermore, reduced upper airway airflow inhibits adequate air movement reducing a BOAS-affected dog’s ability to cool themselves through panting [3, 4, 10]. Cumulatively, dogs with BOAS have a shortened lifespan and increased medical costs compared to non-brachycephalic dogs [1, 2]. Early identification and intervention is recommended to reduce long term consequences of BOAS [11].
Currently veterinarians perform physical examinations, thoracic radiographs, and upper airway examinations to identify dogs with BOAS and attempt to assess disease severity [12]. However, in addition to carrying risks associated with sedation and radiation exposure, this assessment is subjective and not linked to physiologic consequences of BOAS: upper airway inflammation and airflow obstruction, hypoxemia, exercise intolerance, and impaired thermoregulation [3, 4, 10, 13]. As such affected dogs may be missed, or disease severity underestimated, when diagnosis is based on anatomic evaluation alone. A 4-point BOAS grading scheme has been developed that incorporates clinical signs with exercise, and whole body plethysmography, in order to identify affected dogs and stratify them by disease severity [14]. This has been adopted as the Respiratory Function Grading (RFG) scheme, to guide better breeding practices in brachycephalic dogs [15].
The six-minute walk test (6MWT), a standardized measure of athletic performance, has been used to evaluate dogs with respiratory disease [13, 16–18]. The 6MWT has been shown to differentiate normal dogs from those with respiratory diseases, as well as differentiate between dogs with mild and severe BOAS using objective endpoint parameters (i.e., distance walked, heart rate, respiratory rate, rectal temperature, and pulse oximetry) [13, 17, 18]. Collectively, these endpoint parameters are used to detect changes in physical performance capacity and facilitate comparisons in endurance between groups. However, these findings have not been correlated with local upper airway inflammation and or airflow obstruction which contribute to the severity and progression of clinical signs.
Infrared thermography (IRT) provides a safe, non-invasive, and radiation-free method for detecting airflow and inflammation by measuring skin surface temperatures [19, 20]. Skin surface temperature correlates with temperatures in deeper tissues [19, 20]. Though not previously used for evaluation of the upper airway in dogs, IRT has been used to evaluate dogs following exercise and to monitor pain and post-surgical inflammation [21–23]. In people, IRT has been used to assess obstructive sleep apnea [24–26]. The presence and severity of obstructive sleep apnea correlates well with IRT surface temperature measurements over the upper airway [25]. English Bulldogs are animal model of obstructive sleep apnea in humans suggesting possible utility IRT for evaluating dogs with BOAS [7].
The objectives of this study were three-fold. Firstly, to evaluate the utility of the 6MWT and IRT to detect changes in physical performance, upper airway obstruction and or inflammation, and impaired thermoregulation in dogs with BOAS (grade 1–3) compared to healthy brachycephalic (grade 0), and healthy mesaticephalic dogs. Secondly, to determine IRT cutoff values discriminating dogs with BOAS from dogs without BOAS. Thirdly, to evaluate results of the 6MWT and IRT between these groups for correlations between upper airway temperature and performance on a 6MWT. We hypothesize that compared to healthy mesaticephalic and brachycephalic dogs, dogs with BOAS will: 1) show evidence of decreased performance capacity on a 6MWT; 2) display higher upper airway temperatures as assessed by IRT; and 3) upper airway temperatures will correlate with distance walked and rectal temperatures.
Materials and methods
Case selection and criteria
Dogs presenting to Auburn University Wilford and Kate Bailey Small Animal Teaching Hospital (AU-VTH) between July 20, 2021 and February 25, 2022 were prospectively enrolled into one of three groups (Table 1) [14]. Healthy brachycephalic (Brachy) and mesaticephalic (Mesa) dogs were recruited from companion dogs owned by the staff, students, and faculty of the Bailey Small Animal Teaching Hospital at Auburn University. Dogs with clinical BOAS were recruited from the hospital clinical population.
Table 1. BOAS grading scheme.
| Group | Head Conformation | Grade | Respiratory Noise | Inspiratory Effort | Dyspnea/Cyanosis/Syncope |
|---|---|---|---|---|---|
| Mesa | Mesaticephalic | 0 | Pre-6MWT: None | Pre-6MWT: None | Pre-6MWT: None |
| Post-6MWT: None | Post-6MWT: None | Post-6MWT: None | |||
| Brachy | Brachycephalic | 0 | Pre-6MWT: None | Pre-6MWT: None | Pre-6MWT: None |
| Post-6MWT: None | Post-6MWT: None | Post-6MWT: None | |||
| BOAS | Brachycephalic | 1 | Pre-6MWT: None-Mild | Pre-6MWT: None | Pre-6MWT: None |
| Post-6MWT: Mild | Post-6MWT: None-Mild | Post-6MWT: None | |||
| 2 | Pre-6MWT: Mild-Moderate | Pre-6MWT: Mild-Moderate | Pre-6MWT: None | ||
| Post-6MWT: Moderate-Severe | Post-6MWT: Moderate-Severe | Post-6MWT: Mild dyspnea, No cyanosis or syncope | |||
| 3 | Pre-6MWT: Moderate-Severe | Pre-6MWT: Moderate-Severe | Pre-6MWT: Moderate-Severe dyspnea ± cyanosis, inability to exercise | ||
| Post-6MWT: Severe | Post-6MWT: Severe | Post-6MWT: Severe dyspnea, ± cyanosis or syncope |
Adapted brachycephalic obstructive airway syndrome (BOAS) scale [14, 15]. The scale describes the clinical manifestations of upper airway obstruction before and after a six-minute walk test (6MWT). Brachycephalic dogs with a BOAS grade of 0 were enrolled in the Brachy group (i.e., healthy brachycephalic dogs). Brachycephalic dogs with BOAS grade of 1–3 were enrolled in the BOAS group (i.e., clinically affected). Mesaticephalic dogs (Mesa) were also graded to demonstrate lack of clinical evidence of upper airway obstruction (grade 0).
Dogs were included in the Brachy or BOAS groups if they were brachycephalic dogs with BOAS grades 0 and 1–3 respectively (Table 1). Dogs with BOAS were excluded if disease severity was sufficient to prevent safe completion of the study. Dogs were included in the Mesa group if they were a healthy dog with an appropriate craniofacial conformation (mesaticephalic) and no evidence of obstructive upper airway disease. Dogs, regardless group, were excluded from the study if they had previous surgical treatment for BOAS (i.e., stenotic nares, elongated/hyperplastic soft palate, everted laryngeal saccules, laryngeal collapse, or aberrant nasal turbinates), if they had concurrent or historical nasal, respiratory, cardiovascular, or orthopedic disease, or other systemic illness unrelated to BOAS within the preceding 6 months. Dogs were also excluded if their temperament did not allow appropriate image acquisition and measurement of vital parameters (e.g., heart rate, respiratory rate, rectal temperature, and pulse oximetry). Once enrolled, demographic data (age (yrs), breed, sex, body weight (kg), body condition score (BCS; 9-point scale), head conformation (mesaticephalic or brachycephalic), leg length (right front; cm), and duration of clinical signs for BOAS dogs (CS; yrs) were recorded). Leg length was measured from the point of the shoulder to the floor for all dogs. The study protocols were approved by the Auburn University IACUC committee (#2020–3814).
BOAS grading
The criteria for BOAS grading were adapted from previously published studies and the Respiratory Function Grading scheme (Table 1) [14, 15, 27]. BOAS grading was performed by a single individual (JG). Mesaticephalic dogs were also graded using the same criteria to demonstrate lack of clinical evidence of upper airway obstruction (grade 0).
Six-minute walk test
A 6MWT was performed as previously described [16–18]. Briefly, the 6MWT was performed along a dedicated, 150 foot-long, temperature-controlled hallway. Dogs were walked with a harness to avoid temperature changes over the cervical region which could be detected by IRT. Rectal temperature (°C), heart rate (bpm), respiratory rate (brpm), and pulse oximetry (%) data were collected at baseline (pre-6MWT), immediately following (T0), and at 5min- (T5), and 15min-post 6MWT (T15). Dogs were allowed to set their own pace and rest if desired. The distance walked was measured (m) and normalized according to leg length.
Infrared thermographic imaging
All enrolled dogs were acclimatized to a standardized room in our hospital for 30 minutes prior to image collection. The mean temperature in the acclimatization, and IRT rooms as well as the hallway was maintained at [23.3°C (74°F)]. The doors of the rooms remained closed during the data collection period to reduce temperature fluctuations. The dogs’ muzzles and necks were not handled from the time the acclimation period started until after all data was collected to prevent artifactual increases in skin temperature [19]. Dogs had ventral cervical, dorsal, and rostral IRT images collected at baseline, T0, T5, and T15 [25]. A standardized region of interest (ROI) was drawn over each image by a single investigator (JG; Fig 1). The mean temperatures (Tmean) and maximum temperatures (Tmax) within the ROI were calculated and reported. The Tmean calculates the average temperature of all the pixels within the ROI. The Tmax reports the highest temperature pixel within the ROI. For IRT image collection, a Fluke TiX580 camera with a 640 x 480 pixel detector and spectral range of 7.5μm– 14μm was used. The camera was positioned 1.5–2 feet away from the dogs’ necks to standardize distance from the imaged area. An emissivity value of 0.95 was used [19]. Fluke SmartView version 4.3.154.0 software was utilized to analyze the IRT images.
Fig 1. A-C: Infrared thermographic regions of interest (ROI).
Infrared thermographic images depicting ROIs which are outlined in black (arrowhead). A) The ventral ROI was drawn to overly the larynx by centering over the ventral neck between the thoracic inlet and the ramus of the mandible. B) The dorsal ROI was drawn over the bridge of the nose between the medial canthi and extending rostrally to the start of the nasal planum. (C) The rostral ROI is drawn to include the nose with nares. Thermal scales are provided to the right of the images.
Statistical analysis
Statistical analysis was completed using commercial statistical analysis software (Sigma Plot 14.5). Descriptive statistics were applied where appropriate. Normality assessment was performed by the Shapiro-Wilks test. Between group comparisons were made using an ANOVA on RANKS and One-way ANOVA and for non-parametric and parametric data respectively. Non-parametric data are presented as median (IQR) and parametric data as mean ± stdev. Receiver operating characteristic (ROC) curves were used to assess the utility of IRT to discriminate the BOAS group from the non-BOAS groups (i.e., Brachy and Mesa). Spearman rank-order or Pearson correlation test was used to evaluate the association between 6MWT and IRT parameters and between Tmean and Tmax temperatures for each ROI and time point. Extrapolating from data in people, a sample size of 10 dogs per group was calculated to detect a 1°C and 20% difference between groups for IRT and 6MWT respectively using a power of 0.80 and significance level of < .05 [25].
Results
Animals
Thirty-six dogs were prospectively enrolled: Mesa (n = 16), Brachy (n = 10), and BOAS (n = 10). Sixteen breeds were represented (Table 2). Eighteen dogs were male (2 intact; 16 castrated) and 18 were female (2 intact; 16 spayed). There were no significant differences between groups for age, weight, body condition score, or leg length (p ≥.05). Demographic data by group are provided in Table 3. BOAS grades were as follows: Mesa (grade 0; n = 16), Brachy (grade 0; n = 10), and BOAS (grade 1 (n = 6); grade 2 (n = 3) grade 3 (n = 1)). For BOAS dogs, breeds and corresponding BOAS score has been provided in Table 4. For BOAS dogs the median (IQR) duration of clinical signs was 4.88 yrs (1.69–8.50 yrs) (range 1.17–11.8 yrs). Stertor was identified in 10/10 dogs. Stridor was identified in 3/10 BOAS dogs: grade 1 (n = 2) and grade 2 (n = 1). No dog had a history of syncope or had any syncopal events during data collection. All dogs (n = 36) completed the 6MWT successfully completed the study. All dogs were companion animals and per owners did not routinely engage in exercise beyond leash walks.
Table 2. Breeds evaluated.
| Dogs (n) | Breed |
|---|---|
| 5 | French bulldog, Chihuahua mix |
| 4 | English bulldog |
| 3 | Pit Bull, Brussels griffon, mixed breed dog |
| 2 | Boston terrier, Pug, Great Pyrenees |
| 1 | Boxer, pit bull mix, Labrador retriever, terrier mix, boxer mix, golden retriever, dachshund mix |
Number (n) of dogs of various breeds undergoing six-minute walk test (6MWT) and infrared thermography (IRT).
Table 3. Demographic and 6MWT parameters.
| Parameter | Mesa | Brachy | BOAS | P value |
|---|---|---|---|---|
| Age (yrs) | 3.5 (2.6–7.5) | 7.8 (2.7–10.2) | 4.9 (1.7–8.5) | .68 |
| Weight (kg) | 25.0 (7.0–31.9) | 8.3 (5.3–18.6) | 14.1 (10.1–17.5) | .18 |
| BCS (1–9) | 5 (4–5.25) | 5 (5–6) | 6 (5–7.5) | .29 |
| BOAS grade (0–3) | 0 (0–0) | 0 (0–0) | 1 (1–2) | |
| Leg length (Right Front; cm) | 34.3 ± 12.1 | 27.3 ± 11.5 | 23.6 ± 3.9 | .13 |
| 6MWT Distance (m) | 420.4 ± 89.3 | 461.0 ± 60.2 | 379.8 ± 97.6 | .05 |
| Temp (°C) | ||||
| Baseline | 38.6 ± 0.43 | 38.9 ± 0.40 | 38.6 ± 0.70 | .22 |
| T0 | 38.7 ± 0.40 | 39.1 ± 0.40 | 38.8 ± 0.70 | .10 |
| T5 | 38.8 ± 0.40 | 38.9 ± 0.60 | 39.0 ± 0.50 | .51 |
| T15 | 38.7 ± 0.37 | 38.9 ± 0.30 | 38.7 ± 0.60 | .91 |
| Heart Rate (bpm) | ||||
| Baseline | 136.0 (101.0–141.0) | 144 (126–147) | 140.8 (128–151) | .08 |
| T0 | 109.4 ± 19.3a,b | 141.6 ± 14.0b | 145.6 ± 22.5a | < .001* |
| T5 | 106.0 ± 18.0a,b | 133.8 ± 20.4b | 138.8 ± 33.2a | .002* |
| T15 | 100.5 ± 20.6a | 122.6 ± 26.3 | 131.8 ± 30.5a | .01* |
| Respiratory Rate (brpm) | ||||
| Baseline | 36 (27–37) | 32 (24–40) | 28 (24–32) | |
| n = 8/16 | n = 5/10 | n = 5/10 | ||
| T0 | 40 (29–54) | 44 (34–54) | 36 | |
| n = 6/16 | n = 3/10 | (n = 1/10) | ||
| T5 | 28 (24–44) | 38 | 32 | |
| n = 5/16 | (n = 1/10) | n = 1/10 | ||
| T15 | 32 (22–46) | 36 32–36 | 32 (24–34) | |
| n = 7/16 | n = 5/10 | n = 3/10 | ||
| SPO2 (%) | ||||
| Baseline | 97.5 (96.8–100) | 97.5 (97–99) | 98.5 (95.3–100) | .48 |
| T0 | 97.5 (96–99) | 97.5 (97–99.8) | 97.5 (95.3–98.8) | .62 |
| T5 | 98.5 (95.8–100) | 100 (97.5–100) | 97.5 (94.3–98.8) | .16 |
| T15 | 99 (96–100) | 98.5 (96.3–99) | 99 (96–100) | .90 |
Demographic and 6-Minute walk test (6MWT) data are displayed as median (IQR) or mean ± stdev for non-normal and normally distributed data respectively. Statistically significant differences between groups (p <0.05) are denoted by an asterisk (*). Due to panting, too few dogs had discrete respiratory rates to allow for statistical comparisons. Median (IQR) or mean ± stdev are provided where able. The number of dogs (n) with recorded respiratory rates out of the group total are provided. The remaining dogs were panting. 6-minute walk test (6MWT), immediately following 6MWT (T0), 5 minutes post-6MWT (T5), 15 minutes post-6MW (T15), brachycephalic (Brachy), brachycephalic obstructive airway syndrome (BOAS), beats per minute (bpm), breaths per minute (brpm), mesaticephalic (Mesa), oxygen saturation (SPO2), years (yrs).
a: groups are statistically different.
b: groups are statistically different.
Table 4. Breed and BOAS score in dogs with clinical BOAS.
| Breed (n) | BOAS score |
|---|---|
| Boston Terrier (1) | 1 |
| French Bulldog (4) | 1,1,1,2 |
| English Bulldog (3) | 1,2,3 |
| Pug (2) | 1,2 |
Breeds and BOAS scores for dogs with clinical BOAS (scores 1–3). All dogs in the Mesa and Brachy groups had a score of zero.
6-minute walk test
6-minute walk test data by group are provided in Table 3. No significant differences between groups were detected for temperature, SPO2, or normalized distance walked (p ≥.05). Heart rates were statistically higher in Brachy and BOAS dogs compared to Mesa dogs at T0 and T5 and between BOAS and Mesa dogs at T15. No significant differences were detected for heart rate at baseline.
Infrared thermography
Significant differences in Tmean and Tmax temperatures between groups were detected for dorsal and rostral ROIs across all time points (Table 5). No significant differences were detected for ventral ROIs between groups at any time point (p>.05). Dorsal and rostral Tmean and Tmax showed significant strongly positive correlations across all time points and the ventral ROI Tmean and Tmax showed significant moderate-strong positive correlations (Table 6). Receiver operating characteristic curves evaluating the dorsal and rostral Tmean and Tmax between BOAS and non-BOAS (Brachy and Mesa) dogs is provided in Table 7. Temperature cut-offs were selected to optimize sensitivity and specificity. Areas under the curve for the dorsal ROI ranged from 0.85–0.94 (Tmean) and 0.79–0.91 (Tmax). The rostral ROI AUCs ranged from 0.76–0.88 (Tmean) and 0.73–0.96 (Tmax). Dorsal Tmax and Tavg was not available for 1 dog (BOAS) at T15 because the camera angle did not permit accurate drawing of the ROI. Ventral Tmax and Tavg measurements were not available for 1 dog each at T0, T5, T15 respectively. In this case all dogs belonged to the Mesa group. This is because the harness obscured the ROI. All other images and measurements were recorded uneventfully.
Table 5. IRT Tmax and Tmean for ventral, dorsal, and rostral ROIs.
| Mesa | Brachy | BOAS | P value | |
|---|---|---|---|---|
| Tmean | °C | |||
| Ventral | ||||
| Baseline | 33.1 ± 1.8 | 33.3 ± 1.7 | 33.9 ± 1.7 | .52 |
| T0 | 32.3 ± 2.6 | 32.7 ± 2.0 | 34.1 ± 1.8 | .17 |
| T5 | 32.9 (31.5–35.2) | 32.9 (31.6–33.3) | 34.7 (32.8–35.7) | .22 |
| T15 | 33.1 (31.7–35.3) | 33.3 (31.4–34.0) | 33.9 (32.7–35.2) | .54 |
| Dorsal | ||||
| Baseline | 30.47 ± 1.8a | 31.7 ± 2.5b | 35.1 ± 3.2a,b | < .001* |
| T0 | 34.4 (32.1–35.0)a | 35.3 (33.8–36.2) | 37.3 (36.4–37.7)a | .002* |
| T5 | 33.6 (31.8–34.0)a | 34.7 (36.0–36.4) | 37.2 (36.4–37.8)a | < .001* |
| T15 | 32.8 ± 1.2a | 33.6 ± 2.1b | 37.0 ± 1.1a,b | < .001* |
| Rostral | ||||
| Baseline | 24.9 (23.9–26.6)a | 25.4 (24.1–27.7.9) | 30.1 (26.2–35.0)a | .045* |
| T0 | 28.6 ± 3.6a | 30.8 ± 2.6 | 33.5 ± 3.9a | .004* |
| T5 | 27.9 ± 3.2a | 30.3 ± 2.5b | 33.8 ± 3.3a,b | < .001* |
| T15 | 27.3 ± 3.4a | 29.7 ± 3.1b | 33.5 ± 3.4a,b | < .001* |
| Tmax | ||||
| Ventral | ||||
| Baseline | 35.5 ± 1.4 | 36.6 ± 1.7 | 36.7 ± 1.2 | .08 |
| T0 | 35.2 ± 1.6 | 36.6 ± 1.6 | 36.1 ± 1.1 | .05 |
| T5 | 35.8 ± 1.4 | 36.6 ± 1.5 | 36.8 ± 1.0 | .12 |
| T15 | 36.1 ± 1.6 | 36.8 ± 1.4 | 36.6 ± 1.8 | .51 |
| Dorsal | ||||
| Baseline | 34.2 ± 1.9 a | 35.3 ± 2.4 | 37.2 ± 2.2a | .005* |
| T0 | 36.4 ± 1.3 a | 37.3 ± 1.7 | 38.2 ± 2.1 a | .007* |
| T5 | 35.9 ± 1.1 a | 36.8 ± 1.1 | 38.1 ± 1.7 a | < .001* |
| T15 | 35.6 (35.0–36.5) a | 36.6 (36.4–37.2)b | 38.4 (37.8–39.3) a,b | < .001* |
| Rostral | ||||
| Baseline | 29.6 ± 2.5a | 32.0 ± 2.8 | 34.3 ± 4.8a | .025* |
| T0 | 32.9 ± 3.1a | 35.5 ± 1.5 | 36.7 ± 3.4a | < .001* |
| T5 | 32.2 ± 3.0a | 35.0 ± 1.0 | 37.1 ± 2.0a | < .001* |
| T15 | 32.1 ± 2.8a | 34.8 ± 2.0 | 36.7 ± 2.8a | < .001* |
Infrared thermography data are displayed at median (IQR) or mean ± stdev for non-normal and normally distributed data respectively. Data are displayed in ° Celsius. Statistically significant differences between groups (p <0.05) are denoted by an asterisk (*). 6-minute walk test (6MWT), immediately following 6MWT (T0), 5 minutes post-6MWT (T5), 15 minutes post-6MW (T15), brachycephalic (Bracy), brachycephalic obstructive airway syndrome (BOAS), mesaticephalic (Mesa), mean temperature (Tmean), maximum temperature (Tmax).
a: groups are statistically different.
b: groups are statistically different.
Table 6. Correlations between IRT Tmean and Tmax.
| Correlations | r | P values |
|---|---|---|
| Baseline | ||
| Rostral (Tmean vs Tmax) | 0.90 | <0.001* |
| Dorsal (Tmean vs Tmax) | 0.90 | <0.001* |
| Ventral (Tmean vs Tmax) | 0.70 | <0.001* |
| T0 | ||
| Rostral (Tmean vs Tmax) | 0.93 | <0.001* |
| Dorsal (Tmean vs Tmax) | 0.88 | <0.001* |
| Ventral (Tmean vs Tmax) | 0.77 | <0.001* |
| T5 | ||
| Rostral (Tmean vs Tmax) | 0.85 | <0.001* |
| Dorsal (Tmean vs Tmax) | 0.87 | <0.001* |
| Ventral (Tmean vs Tmax) | 0.54 | <0.001* |
| T15 | ||
| Rostral (Tmean vs Tmax) | 0.88 | <0.001* |
| Dorsal (Tmean vs Tmax) | 0.85 | <0.001* |
| Ventral (Tmean vs Tmax) | 0.76 | <0.001* |
Pearson correlation coefficients and p values are displayed for Tmean vs Tmax IRT readings for each ROI. Dorsal and rostral ROIs showed significant strongly positive correlations between Tmean and Tmax. The ventral ROI showed significant moderate to strong positive correlations between Tmean and Tmax. A p < .05 is considered significant (*). 6-minute walk test (6MWT), immediately following 6MWT (T0), 5 minutes post-6MWT (T5), 15 minutes post-6MW (T15), mean temperature (Tmean), maximum temperature (Tmax).
Table 7. Receiver operating characteristic analysis of BOAS and non-BOAS dogs.
| Receiver Operating Characteristic Analysis | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| AUC | Cutoff (>°C) | SE | 95% CI | SP | 95% CI | LR+ | LR- | P value | ||
| Dorsal | ||||||||||
| Tmean | Baseline | 0.85 | 35.1 | 0.70 | 0.35–0.93 | 1.0 | 0.87–1 | - | 0.3 | .001 |
| T0 | 0.85 | 35.7 | 0.90 | 0.56–1 | 0.80 | 0.61–0.93 | 4.7 | 0.12 | .001 | |
| T5 | 0.90 | 35.6 | 0.88 | 0.56–1 | 0.90 | 0.7–0.98 | 7.8 | 0.11 | .0003 | |
| T15 | 0.94 | 35.3 | 1.0 | 0.66–1 | 0.92 | 0.75–0.99 | 13.0 | 0 | < .0001 | |
| Tmax | Baseline | 0.81 | 35.7 | 0.80 | 0.44–0.94 | 0.73 | 0.52–0.88 | 3.7 | 0.33 | .004 |
| T0 | 0.79 | 37.7 | 0.80 | 0.44–0.97 | 0.73 | 0.52–0.88 | 3.7 | 0.33 | .0007 | |
| T5 | 0.86 | 37.9 | 0.80 | 0.44–0.97 | 0.92 | 0.8–1 | 4.8 | 0.05 | .0007 | |
| T15 | 0.91 | 37.6 | 0.88 | 0.52–1 | 0.92 | 0.75–0.99 | 4.1 | 0.1 | < .0001 | |
| Rostral | ||||||||||
| Tmean | Baseline | 0.76 | 27.3 | 0.7 | 0.35–0.93 | 0.81 | 0.61–0.93 | 3.6 | 0.37 | .016 |
| T0 | 0.84 | 31.7 | 0.9 | 0.56–1 | 0.85 | 0.65–0.96 | 5.9 | 0.11 | .001 | |
| T5 | 0.88 | 32.3 | 0.9 | 0.56–1 | 0.92 | 0.75–1 | 11.7 | 0.10 | .0003 | |
| T15 | 0.87 | 31.8 | 0.9 | 0.56–1 | 0.85 | 0.65–0.96 | 5.9 | 0.11 | .0006 | |
| Tmax | Baseline | 0.73 | 33.5 | 0.7 | 0.35–0.93 | 0.88 | 0.7–0.98 | 6.1 | 0.34 | .03 |
| T0 | 0.84 | 31.2 | 0.9 | 0.56–1 | 0.81 | 0.61–0.93 | 4.7 | 0.12 | .002 | |
| T5 | 0.92 | 36.4 | 0.9 | 0.6–1 | 1.0 | 0.9–1 | - | 0.10 | .0001 | |
| T15 | 0.96 | 36.8 | 1.0 | 0.7–1 | 0.85 | 0.7–1 | 6.5 | 0 | < .0001 | |
Receiver operating characteristic (ROC) curve data for Dorsal and Rostral ROIs (Tmax and Tmean) used to discriminate between dogs with clinical BOAS and those without BOAS (Brachy and Mesa). Area under the curve (AUC), Sensitivity (SE), Confidence interval (CI), Specificity (SP), Positive likelihood ratio (LR+), Negative likelihood ratio (LR-). Temperature cutoffs were selected for optimal sensitivity and specificity. 6-minute walk test (6MWT), immediately following 6MWT (T0), 5 minutes post-6MWT (T5), 15 minutes post-6MW (T15), mean temperature (Tmean), maximum temperature (Tmax).
Correlations between IRT and 6MWT
No significant correlations were identified between distance walked for either ventral, dorsal or rostral Tmean at any time point (p≥.05). Significant weakly negative correlations were identified between dorsal Tmax and normalized distance walked at T5 (p = .04; r = -0.337) and T15 (p = .03; r = -0.363). Significant moderately positive correlations were identified between dorsal Tmean and rectal temperatures at baseline (p = .005; r = 0.435), T0 (p = .003; r = 0.479), and T5 (p = .003; r = 0.475). Additionally, significant weak to moderately positive correlations were identified between rostral Tmean and rectal temperatures at baseline (p = .03; r = 0.358), and T0 (p = .006; r = 0.446). Significant weak-moderately positive correlations were identified between dorsal Tmax and rectal temperatures at baseline (p = .01; r = 0.398), T0 (p = .002; r = 0.494) and T5 (p = .04; r = 0.334). Significant moderately positive correlations were identified between rostral Tmax and rectal temperatures at baseline (p = .002; r = 0.484) and T0 (p < .001; r = 0.539).
Discussion
In this pilot study, the combination of a 6MWT and IRT detected evidence suggesting decreased physical performance capacity (i.e., exercise intolerance) and increased dorsal and rostral ROI temperatures in dogs with clinical evidence of BOAS compared to healthy brachycephalic and healthy mesaticephalic dogs. Importantly, dorsal and rostral Tmean and Tmax were able to discriminate BOAS dogs from dogs without BOAS (i.e., Brachy and Mesa groups) across multiple time points. Upper respiratory tract temperatures demonstrated moderate positive correlations with rectal temperatures and weak negative correlations with normalized distance walked. This pilot study demonstrates the potential value of utilizing the 6MWT and IRT, paired with evaluation of BOAS-compatible clinical signs, to diagnose BOAS and its physiologic consequences in a non-invasive, low risk, and objective manner.
Brachycephalic dogs are increasingly popular despite well documented problems associated with their conformation [28]. Early identification of affected dogs is recommended to reduce long-term complications [11]. This makes clinical recognition of affected dogs of clear clinical importance. Currently the criterion standard for diagnosis involves a physical and upper airway examination with thoracic radiography to identify the characteristic anatomic changes [3]. In addition to carrying risks associated with sedation and or anesthesia (i.e., upper airway examination), these do not identify the important down-stream physiologic consequences of BOAS [3, 5, 10, 13]. Identifying the pathologic consequences of BOAS may significantly impact management recommendations in addition to promoting early identification of affected dogs through low risk clinical screening.
For our pilot study, dogs were evaluated for CS consistent with BOAS adapted from a previously validated scoring system [8, 14, 15, 27]. This method strongly correlated clinical evidence of BOAS with objective evidence of upper airway obstruction using whole body plethysmography and predicted BOAS status with 94–97% accuracy [13, 14, 16, 27]. Similar grading schemes are used to promote responsible breeding practices in brachycephalic dogs (i.e., Pugs, French Bulldogs, and English Bulldogs) [15]. As such, and for the purposes of this study, group assignments were considered appropriate despite the absence of an upper airway examination. In our study, most dogs were considered to have mild BOAS based clinical grade (grade 1: n = 6). Between group comparisons demonstrated no differences in demographic parameters including BCS and leg length which may influence the results of IRT and the 6MWT [29]. Despite the lack of significant differences in leg length, a normalized distance walked was reported to mitigate even minor differences [16].
The 6MWT allows comparison of physical performance between groups following standardized submaximal exercise [16–18]. In previous studies in dogs, the 6MWT differentiated normal dogs from those with respiratory diseases. Dogs with respiratory disease walked significantly less distance during the prescribed six minutes than did normal dogs [13, 17, 18]. The 6MWT has also been utilized to compare English Bulldogs with severe clinical signs of BOAS with those who were only mildly affected. The more severely affected dogs walked a significantly shorter distance during the 6MWT, and took for long for recovery (i.e., heart rate, respiratory rate, upper respiratory noise, and body temperature returning to baseline) [13]. Though no differences in distance walked were detected between groups, the dogs in our study showed differences in other parameters of physical performance. In our study no differences in heart rate between groups was detected at baseline. However, significant differences were detected between groups following exercise [30]. Interestingly, all groups showed subjective decreased heart rate compared to baseline at T5 and T15. The Mesa and Brachy but not BOAS groups showed decreased heartrate compared to baseline at T0. This initial change in heartrate was attributed to decreased excitement compared to the start of the test. Though the heart rates for all groups apparently decreased, they still showed between group difference. At T0 and T5 both the BOAS and Brachy groups demonstrated increased heart rates compared to the Mesa group. Interestingly, at T15 the BOAS group maintained statistically increased heart rates compared to the Meso group while the Brachy group did not. These differences in markers of endurance may suggest that not only do BOAS dogs have evidence of decreased athletic performance compared to mesaticephalic dogs, but they also take longer to recover after exertion compared to their healthy brachycephalic counterparts. No other differences in 6MWT parameters were identified; respiratory rate, and SPO2. The lack of significant differences between groups for other 6MWT parameters may reflect type 2 error as our population was biased toward less clinically affected dogs. Ultimately, larger longitudinal studies are required before the clinical significance of these finding can be determined.
The English Bulldog is an animal model of obstructive sleep apnea (OSA) in humans which, like BOAS, is characterized by upper respiratory tract inflammation and airflow limitation [5, 12, 25]. In humans, the presence and severity of OSA correlates well with IRT surface temperature measurements [25]. In our study, between group comparisons identified significant differences in Tmax and Tmean for the dorsal and rostral but not ventral ROIs. While no significant differences were detected between the Brachy and Mesa groups at any time point, significant differences were detected between both the BOAS and Mesa and BOAS and Brachy groups. These preliminary findings may suggest that clinical signs of disease, based on our scoring system, were associated with objective markers of inflammation and or airflow limitation on IRT, rather than brachycephalic head conformation alone. The reason for the lack of statistical differences using the ventral ROI is unknown as this is a common site of evaluation on people, however, redundant skin folds and differences in haircoat may be considered as a possible barrier due to increased distance from the airway compared to people [25].
Increased dorsal and rostral Tmean and Tmax was found for the BOAS group compared to the Mesa group at all time points. Baseline dorsal Tmean was also higher in the BOAS group compared to the Brachy group. Increased temperatures at baseline may suggest underlying inflammation as in human studies of OSA [25]. However, as dogs rely upon evaporative cooling, increased temperatures may reflect a failure of normal heat transfer and impaired thermoregulation because of obstructed airflow. This may be supported by comparing the upper airway temperatures between the BOAS and Brachy groups after exercise. The baseline rostral Tmean and dorsal and rostral Tmax were not significantly different between the BOAS and Brachy groups. However, the BOAS group had increased post-6MWT temperatures compared to the Brachy group at multiple timepoints (Table 4). The dorsal and rostral ROI temperatures were also correlated with rectal temperature. This may support BOAS as a contributor to impaired thermoregulation however concurrent contribution of inflammation to this finding is unknown as inflammation has also been detected in dogs with BOAS [5]. Additional, larger scale studies in this area are needed. Importantly however, these between group differences were not observed on rectal temperature during 6MWT at any timepoint. Therefore, a combination local (i.e., IRT) and systemic (6MWT) evaluation may be useful to evaluate dogs particularly those with low grade disease. Interestingly, dogs in the Brachy group showed no significant differences in ROI temperatures compared to the Mesa group which may imply that the afore mentioned differences reflect clinical BOAS rather than a brachycephalic head conformation in the absence of disease. Receiver operating characteristic analysis demonstrated good-excellent discrimination between the BOAS group and those without BOAS (i.e., Brachy and Meso groups) for both Tmean and Tmax. The findings of this pilot study may suggest that IRT with 6MWT could provide a method to screen at risk dogs to help with early, low risk detection of BOAS.
In keeping with our hypothesis, significant moderately positive correlations were identified between ROI temperatures and rectal temperature. This is in keeping with previous studies where IRT measurements at the extremities (e.g. pinna and eye) correlated with rectal temperatures and, as stated above, may reflect impaired thermoregulation [22]. Additionally, significant weakly negative correlations were found between ROI temperatures and distance walked on 6MWT. The weak correlations between IRT and distance walked may reflect the low-grade clinical signs demonstrated by the dogs in the BOAS group.
In our study both Tmax and Tmean are reported. The Tmean is commonly reported in the human and veterinary literature and often selected because it is less affected by measurement noise [22, 23, 31–33]. However, this limits automation as it requires carefully hand drawing ROIs to avoid including unintended pixels. The Tmax reduces operator error and allows for ROI automation because only the warmest pixels are counted [32]. Consistent with studies in people, in our study Tmax and Tmean were correlated [32, 33]. The Tmean and Tmax were significantly and strongly correlated for dorsal and rostral ROIs and moderately-strongly positively correlated for ventral ROIs [32, 33]. However, studies evaluating true agreement between these two measurements, as well as intra- and inter-observer variability, are lacking in the veterinary literature and it is therefore unknown if one could be performed in leu of the other. As such, for this pilot study it was elected to report both Tmean and Tmax measurements [32, 33]. Additional studies focused on these important clinical practicalities are needed before this technique can be widely adopted.
There are a few limitations to this study. Firstly, this study is in a relatively small number of dogs. Though we met our a-priori sample size calculation, additional larger scale studies are recommended including dogs with more significant clinical signs of BOAS. Further, these results have not been evaluated against more conventional methods of diagnosis (e.g., upper airway examination). Though clinical assessment and scoring of brachycephalic dogs based on CS has been validated in dogs [13, 14], as of yet no conclusions can be drawn between the findings on IRT and specific anatomic abnormalities. As such, dogs would still need to undergo conventional evaluation prior to surgical correction. However, IRT and 6MWT could be used as a low risk, minimally invasive screening tool to identify at risk dogs which may benefit from further evaluation. This could be important in cases of early disease or where clients are reluctant to consider anatomic assessment due to cost, risk, or acceptance of clinical signs of BOAS as “normal” leading to delayed recognition and intervention [34]. This technique may also be useful in evaluating non-surgical interventions such as weight loss. An additional limitation is that despite no statistically significant differences in demographic parameters, dogs were not age-, breed-, or weight-matched between groups. Because of the small sample size of this pilot study such comparisons were not performed and breed associated affects cannot be excluded. The ROIs were also hand-drawn which may inadvertently introduce error. We attempted to mitigate this by allowing only one individual (JG) to perform the ROI tracing. We also included evaluation of Tmax measurements which are less dependent of consistent drawing of ROIs, though intra-observer variability was not performed. Though the technique was generally easy to perform, the ventral ROI was obscured for 1 time point each in 3 Mesa dogs and the dorsal ROI at 1 timepoint in a BOAS dog. The ventral ROI is not likely to be useful give the limitations in data collection as well as the lack of between group differences and differences in neck folds and haircoat. Though dogs had minimal hair covering for the dorsal ROI differences in skin folds may impact IRT results. However, the rostral ROI was successfully collected in all dogs at all timepoints and is less likely to be affected by haircoat, skin color, and skin thickness (e.g., skin folds). Though further studies are needed, this ROI may be the most practical for clinical use. Finally, while BOAS grading and 6MWT were performed by the same individual (JG) IRT images were collected by multiple investigators (MG, TLF, SCP).
Conclusion
This pilot study demonstrates the potential utility of pairing the 6MWT and IRT with evaluation of CS as screening tool to identify affected dogs with BOAS. This may aid in early clinical diagnosis and improve clinical recommendations by providing a minimally invasive, safe, and objective method of identification of dogs by pairing clinical signs with the physiologic consequences of BOAS: upper airway inflammation and or airflow limitation, exercise intolerance, and impaired thermoregulation.
Supporting information
(XLSX)
Data Availability
All relevant data are within the paper and its Supporting Information files.
Funding Statement
The authors received no funding for this work.
References
- 1.O’Neill DG, Jackson C, Guy JH, Church DB, McGreevy PD, Thomson PC, et al. Epidemiological associations between brachycephaly and upper respiratory tract disorders in dogs attending veterinary practices in England. Canine Genetics and Epidemiology. 2015;2(1):1–10. doi: 10.1186/s40575-015-0023-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Chaters G, Trees AJ, Laing G. Higher insurance premiums revealed for popular brachycephalic BREEDs. [Google Scholar]
- 3.Packer RM, Tivers MS. Strategies for the management and prevention of conformation-related respiratory disorders in brachycephalic dogs. Veterinary Medicine: Research and Reports. 2015;6:219. doi: 10.2147/VMRR.S60475 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Packer RM, Hendricks A, Tivers MS, Burn CC. Impact of facial conformation on canine health: brachycephalic obstructive airway syndrome. PloS one. 2015;10(10):e0137496. doi: 10.1371/journal.pone.0137496 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Rancan L, Romussi S, Garcia P, Albertini M, Vara E, de la Muela MS. Assessment of circulating concentrations of proinflammatory and anti-inflammatory cytokines and nitric oxide in dogs with brachycephalic airway obstruction syndrome. Am J Vet Res. 2013;74(1):155–60. doi: 10.2460/ajvr.74.1.155 [DOI] [PubMed] [Google Scholar]
- 6.Reeve E, Sutton D, Friend E, Warren‐Smith C. Documenting the prevalence of hiatal hernia and oesophageal abnormalities in brachycephalic dogs using fluoroscopy. Journal of Small Animal Practice. 2017;58(12):703–8. doi: 10.1111/jsap.12734 [DOI] [PubMed] [Google Scholar]
- 7.Hendricks J, Kline L, Kovalski R, O’Brien J, Morrison A, Pack A. The English bulldog: a natural model of sleep-disordered breathing. Journal of Applied physiology. 1987;63(4):1344–50. doi: 10.1152/jappl.1987.63.4.1344 [DOI] [PubMed] [Google Scholar]
- 8.Liu N-C, Troconis EL, Kalmar L, Price DJ, Wright HE, Adams VJ, et al. Conformational risk factors of brachycephalic obstructive airway syndrome (BOAS) in pugs, French bulldogs, and bulldogs. PloS one. 2017;12(8):e0181928. doi: 10.1371/journal.pone.0181928 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Grobman M, Masseau I, Reinero C. Aerodigestive disorders in dogs evaluated for cough using respiratory fluoroscopy and videofluoroscopic swallow studies. Vet J. 2019;251:105344. doi: 10.1016/j.tvjl.2019.105344 [DOI] [PubMed] [Google Scholar]
- 10.Davis MS, Cummings SL, Payton ME. Effect of brachycephaly and body condition score on respiratory thermoregulation of healthy dogs. Journal of the American Veterinary Medical Association. 2017;251(10):1160–5. doi: 10.2460/javma.251.10.1160 [DOI] [PubMed] [Google Scholar]
- 11.Meola SD. Brachycephalic airway syndrome. Top Companion Anim Med 2013;28(3):91–6. doi: 10.1053/j.tcam.2013.06.004 [DOI] [PubMed] [Google Scholar]
- 12.Lodato DL, Hedlund CS. Brachycephalic airway syndrome: pathophysiology and diagnosis. Compend Contin Educ Vet. 2012;34(7):E3. [PubMed] [Google Scholar]
- 13.Lilja-Maula L, Lappalainen AK, Hyytiäinen HK, Kuusela E, Kaimio M, Schildt K, et al. Comparison of submaximal exercise test results and severity of brachycephalic obstructive airway syndrome in English bulldogs. The Veterinary Journal. 2017;219:22–6. doi: 10.1016/j.tvjl.2016.11.019 [DOI] [PubMed] [Google Scholar]
- 14.Liu N-C, Sargan DR, Adams VJ, Ladlow JF. Characterisation of brachycephalic obstructive airway syndrome in French bulldogs using whole-body barometric plethysmography. PLoS One. 2015;10(6):e0130741. doi: 10.1371/journal.pone.0130741 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Ladlow J. Brachycephalic obstructive airway syndrome: guide to the respiratory functional grading scheme. In Practice. 2021;43(10):548–55. [Google Scholar]
- 16.Swimmer RA, Rozanski EA. Evaluation of the 6‐minute walk test in pet dogs. Journal of veterinary internal medicine. 2011;25(2):405–6. doi: 10.1111/j.1939-1676.2011.0689.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Villedieu E, Rutherford L, Ter Haar G. Brachycephalic obstructive airway surgery outcome assessment using the 6-minute walk test: a pilot study. J Small Anim Pract. 2019;60(2):132–5. doi: 10.1111/jsap.12942 [DOI] [PubMed] [Google Scholar]
- 18.Lilja‐Maula L, Laurila H, Syrjä P, Lappalainen A, Krafft E, Clercx C, et al. Long‐term outcome and use of 6‐minute walk test in West Highland White Terriers with idiopathic pulmonary fibrosis. Journal of Veterinary Internal Medicine. 2014;28(2):379–85. doi: 10.1111/jvim.12281 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Vainionpää M. Thermographic imaging in cats and dogs: usability as a clinical method. 2014. [Google Scholar]
- 20.Usamentiaga R, Venegas P, Guerediaga J, Vega L, Molleda J, Bulnes FG. Infrared thermography for temperature measurement and non-destructive testing. Sensors. 2014;14(7):12305–48. doi: 10.3390/s140712305 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Soroko M, Górniak W, Howell K, Zielińska P, Dudek K, Eberhardt M, et al. Changes in Body Surface Temperature Associated with High-Speed Treadmill Exercise in Beagle Dogs Measured by Infrared Thermography. Animals. 2021;11(10):2982. doi: 10.3390/ani11102982 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Zanghi BM. Eye and ear temperature using infrared thermography are related to rectal temperature in dogs at rest or with exercise. Frontiers in Veterinary Science. 2016;3:111. doi: 10.3389/fvets.2016.00111 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Casas-Alvarado A, Mota-Rojas D, Hernández-Ávalos I, Mora-Medina P, Olmos-Hernández A, Verduzco-Mendoza A, et al. Advances in infrared thermography: Surgical aspects, vascular changes, and pain monitoring in veterinary medicine. J Therm Biol. 2020;92:102664. doi: 10.1016/j.jtherbio.2020.102664 [DOI] [PubMed] [Google Scholar]
- 24.de Andrade Fernandes A, dos Santos Amorim PR, Brito CJ, Sillero-Quintana M, Marins JCB. Regional skin temperature response to moderate aerobic exercise measured by infrared thermography. Asian journal of sports medicine. 2016;7(1). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Jarnalo M, Vardasca R, Mendes JG, Drummond M. Antero-cervical thermophysiological characterization of obstructive sleep apnea patients. Sleep Breath. 2018;22(4):1111–6. doi: 10.1007/s11325-018-1644-4 [DOI] [PubMed] [Google Scholar]
- 26.Formenti D, Ludwig N, Gargano M, Gondola M, Dellerma N, Caumo A, et al. Thermal imaging of exercise-associated skin temperature changes in trained and untrained female subjects. Annals of biomedical engineering. 2013;41(4):863–71. doi: 10.1007/s10439-012-0718-x [DOI] [PubMed] [Google Scholar]
- 27.Liu NC, Adams VJ, Kalmar L, Ladlow JF, Sargan DR. Whole-Body Barometric Plethysmography Characterizes Upper Airway Obstruction in 3 Brachycephalic Breeds of Dogs. J Vet Intern Med. 2016;30(3):853–65. Epub 20160509. PubMed Central PMCID: PMC4913582. doi: 10.1111/jvim.13933 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Reisen J. Most Popular Dog Breeds of 2021 https://www.akc.org/expert-advice/dog-breeds/most-popular-dog-breeds-of-2021/: The American Kennel Club 2021 [updated March 15, 2022; cited 2022 6/6/22]. [Google Scholar]
- 29.Salamunes ACC, Stadnik AMW, Neves EB. The effect of body fat percentage and body fat distribution on skin surface temperature with infrared thermography. J Therm Biol. 2017;66:1–9. Epub 20170318. doi: 10.1016/j.jtherbio.2017.03.006 [DOI] [PubMed] [Google Scholar]
- 30.Doxey S, Boswood A. Differences between breeds of dog in a measure of heart rate variability. Veterinary Record. 2004;154(23):713–7. doi: 10.1136/vr.154.23.713 [DOI] [PubMed] [Google Scholar]
- 31.Rekant SI, Lyons MA, Pacheco JM, Arzt J, Rodriguez LL. Veterinary applications of infrared thermography. American Journal of Veterinary Research. 2016;77(1):98–107. doi: 10.2460/ajvr.77.1.98 [DOI] [PubMed] [Google Scholar]
- 32.Formenti D, Ludwig N, Rossi A, Trecroci A, Alberti G, Gargano M, et al. Skin temperature evaluation by infrared thermography: Comparison of two image analysis methods during the nonsteady state induced by physical exercise. Infrared Physics & Technology. 2017;81:32–40. 10.1016/j.infrared.2016.12.009. [DOI] [Google Scholar]
- 33.Ludwig N, Formenti D, Gargano M, Alberti G. Skin temperature evaluation by infrared thermography: Comparison of image analysis methods. Infrared Physics & Technology. 2014;62:1–6. 10.1016/j.infrared.2013.09.011. [DOI] [Google Scholar]
- 34.Packer RM, O’Neill DG, Fletcher F, Farnworth MJ. Great expectations, inconvenient truths, and the paradoxes of the dog-owner relationship for owners of brachycephalic dogs. PLoS One. 2019;14(7):e0219918. doi: 10.1371/journal.pone.0219918 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
(XLSX)
Data Availability Statement
All relevant data are within the paper and its Supporting Information files.