Development of a detailed canine gait analysis method for evaluating harnesses: A pilot study

Zsófia Pálya; Kristóf Rácz; Gergely Nagymáté; Rita M Kiss

doi:10.1371/journal.pone.0264299

. 2022 Mar 9;17(3):e0264299. doi: 10.1371/journal.pone.0264299

Development of a detailed canine gait analysis method for evaluating harnesses: A pilot study

Zsófia Pálya ¹, Kristóf Rácz ¹, Gergely Nagymáté ¹, Rita M Kiss ^1,^*

Editor: Ewa Tomaszewska²

PMCID: PMC8906618 PMID: 35263359

Abstract

Dog harnesses are becoming more popular, with their large variety stemming from the idea that different dogs and scenarios require different types of harnesses. While their benefits over collars are self-explanatory, there is a lack of research on their effect on gait, and even the existing studies examine only a limited set of parameters. The goal of present study was to establish a method capable of quantifying canine gait in detail. Based on 3D motion capture, the developed method allows for the examination of 18 joint angles and 35 spatio-temporal parameters throughout multiple gait cycles, and can be used to analyze canine movement in detail in any kind of scenario (e.g. comparing healthy and lame dogs, or measuring the effect of training). The method is presented through the measurement of how different harnesses affect walking kinematics compared to free (unleashed) movements. Four dogs with varying body sizes and breeds and multiple types of harnesses were included. Marker data was filtered using a zero-lag 6th order Butterworth-filter with a cutoff frequency of 20 Hz. The normality of the spatio-temporal and joint range of motion parameters was tested using the Anderson-Darling test (p = 0.05), with most parameters passing in 60+% of test cases. Swing time and range of motion of the sagittal aspect of spinal angle at T1 vertebrae failed more regularly, both resulting from the measurement setup rather than the actual parameters being not normally distributed. Two-sample Kolmogorov-Smirnov tests (p = 0.05) were used to compare each parameter’s distribution between cases, showing that most parameters are significantly altered by the harnesses in about 2/3rd of the cases. Based on the results, there’s no absolute superior harness, however, it is possible to select the best fit for a specific dog and application, justifying their large variety.

Introduction

Gait analysis is a well-established and objective technique to assess normal and abnormal gait accurately, identifying characteristic features of specific gait abnormalities. Besides the wide range of application in human motion analysis (e.g. medicine, sport, rehabilitation), gait evaluation can also be used in veterinary medicine, e.g. the diagnostics of lameness in horses or dogs [1–3]. The measurement techniques within gait analysis can be divided into two primary groups: kinetic and kinematic. Kinetic measurements mostly focus on the forces acting between the foot and the ground (ground reaction forces), measured with force platforms or a baropodometric system, while kinematic measurements record the position and orientation of body segments or bony landmarks, from which the linear and angular velocities, accelerations and the joint angles are determined [4]. With precise measurement techniques, it is possible to quantify gait, so that different groups or populations can be compared numerically.

Veterinary science started to apply the techniques of human motion analysis to study the lameness of horses [1]. In canine science, quantitative gait analysis appeared in the late 20th century [5–7]: the first application was the examination of the resultant truncal and limb transmission force; secondly, joint-specific musculoskeletal functions were analysed with 2-dimensional gait analysis [8]. Nowadays, three-dimensional motion capture is the most common method for examining both human and animal movement patterns. Besides lameness, these various methods may be suitable for analysing different pathologies, such as gait patterns before and after therapeutic treatments or a distinct environmental effect on gait [8–12].

Instead of the regular neck collar, harnesses are becoming more popular, partly due to the emerging canine sports activities, and partly because it can provide better control over dogs, even with the presence of behavioural issues. There are many styles of body harnesses, which can generally be classified as restrictive or non-restrictive. Non-restrictive harnesses feature a Y-shaped chest strap across the body above the scapula, while restrictive harnesses have a strap coming across the chest, crossing the body at, or below the scapula [13]. The large variety of harnesses is the result of an empirical idea that different harnesses are better for certain dogs or scenarios than others. While the benefits of a harness over the neck collar seem self-explanatory, the downsides are more subtle. Most owners and trainers do not observe the gait alternations caused by the harness. These may contribute to long-term repetitive strain, leading to or predisposing an injury [9]. Based on Carr et al., during irregular activates (i.e. training or competing), an objective gait analysis should be performed with the dog wearing the harness to identify any gait alterations [14]. However, according to Blake S. et al., only three full papers were written about examining the biomechanical effects of harness and neck collar use in dogs until September 2019 [15]. Lafuente et al. studied the differences between restrictive and non-restrictive harnesses on shoulder extension, and both harnesses showed significantly decreased joint angle range of motion (ROM) [13]. These conclusions agree with those by Carr et al. in a conference abstract paper, based on examining five different harnesses using a pressure sensitive walkway [14]. Peham et al. investigated the movement of the spine [16] and the pressure distribution under three different types of harnesses designed for guided dogs [17]. They found that spinal motion changed significantly, and calculated forces were greater under the trunk strap. Furthermore, they concluded that there were measurable differences between three types of harnesses. One more study published in March of 2020 investigated the truncal motion of dogs in service vests [18], where they found that the vest has significantly changed the truncal motion of the animal.

A fundamental limitation of previous studies is the low complexity of analysis. They rarely examine more than a few kinematic parameters (e.g. stance time percentage, stride length, and step length [14]) or only calculate minimum, maximum and ROM for joint angles at most (e.g. maximum shoulder extension [13]), instead of describing the joint angle over the full gait cycle. This pilot study’s first goal is to establish a measurement method capable of quantifying canine gait in it’s great complexity, to support long term research, be it the effects of different harnesses have on movement, or any other research related to dog motion. This measurement method should be able to determine spatio-temporal parameters of all four limbs (35 scalar parameters) and the joint angles of the major joints and spinal angles (18 joint angles throughout the whole gait cycle, as well as scalar ROM parameters) captured through multiple gait cycles. We hypothesised that the method will be suitable for all sizes of dogs and different harness types.

Methods

Study design

Four dogs could be recruited for this pilot study whose owners agreed on training their dogs for treadmill walking. This included small, medium and large body sizes with varying breeds, to test the suitability of the method for different circumstances. During the pre-measurements period, the owners trained the dogs to walk on a treadmill for 4–6 weeks (depending on progress), one or two 15 minute sessions per week.

On measurement day, the dog and its owner were provided a 20 minutes accommodation session to become familiar with the treadmill and the environment [2]. Each owner determined the belt velocity during the habituation process so that it seemed comfortable for their dogs. This selected speed was then kept constant for that dog throughout the recordings. Gait patterns were identified by eye by an expert (detailed description of each gait pattern can be found in [19]). The measurements were performed on a ProFitness L150 treadmill (Argos, Milton Keynes, UK) which has a 123 cm long and 40 cm wide belt and has a maximum belt velocity of 12 km/h. During the measurements, the dog’s owner was squatting in front of the treadmill and periodically praised the dog with treats to maintain a continuous walk (Fig 1). The processed section of the trial was a selected homogeneous gait section between treats, usually containing 20 to 30 gait cycles. The movement patterns of each dog were recorded by applying three different types of measurement conditions. Firstly, the reference motion was captured without the dogs wearing any harness. Next, the motion was recorded with the different harnesses without a leash attached, and lastly while wearing each harness with a leash attached. The number of recorded sessions was determined by the number of available harnesses, summarised in Table 1. The order of measurements was the same for all dogs.

Table 1. Participating dogs and used harnesses.

Gait patterns were identified by eye by an expert. Detailed description of each gait pattern can be found in [19].

ID	Breed	Weight (kg)	Harnesses studied	Gait pattern	Treadmill velocity (m/s)
Dog 1	Bullterrier	26	Julius-K9^®power, Julius-K9^®IDC, Julius-K9^®Duo-Flex	pace	0.91
Dog 2	Bullterrier	16	Julius-K9^®power, Julius-K9^®IDC, Julius-K9^®Duo-Flex	amble	0.92
Dog 3	Yorkshire terrier	3	Julius-K9^®power, Julius-K9^®IDC	walk	0.8
Dog 4	Beagle-labrador mixed breed	26	Julius-K9^®power, Julius-K9^®IDC, Julius-K9^®Duo-Flex, Fressnapf own-branded	amble	1.05

Open in a new tab

Dog studied

Four clinically healthy short-haired dogs aged between 3 and 5 years participated in the study, whose anthropometric data is summarised in Table 1. All dogs participated in the measurement with their owners, who have given their written consent to participate in the experiment after they were informed verbally about all aspects (e.g. hazardous situations, instrumentation, use of media contents, etc.). The study was approved by the National Science and Research Ethics Committee (Hungary) (21/2015).

Harnesses studied

An important consideration was to select both restrictive and non-restrictive (Y-type) harnesses into our study, which are commercially available to most owners in the country and manufactured in several sizes. In this research, three harness types were studied which were manufactured by Julius-K9^®and one which was provided by the dog’s owner (only in case of Dog 4). Two of them were restrictive harnesses, namely the Julius-K9^®power harness (Fig 2c) and the Julius-K9^®IDC harness (Fig 2b). The Julius-K9^®Duo-Flex harness (Fig 2a) was a non-restrictive one. These harnesses were manufactured specifically for this research without the light-reflexive materials, which could interfere with motion capture measurements (this did not affect the mechanical properties of the harnesses). In the case of Dog 4, a generic non-restrictive harness was also included (Fig 2d). This harness was a pet store (Fressnapf GmbH, Krefeld, Germany) bought own-branded product. Since it was made with light-reflective patterns, the reflective surfaces was covered with adhesive tapes during the measurements. No regular neck collar was included in the study, as it was determined to not interfere with marker placement as harnesses could, and therefore the developed measurement should be also applicable to collars without issues. All the dogs were previously accustomed to wearing the harnesses. Table 1 also shows which dog was measured with which harnesses. A total of 28 trials was recorded, including all free, harness only, and leashed trials.

Fig 2 — a) Julius-K9^®Duo-Flex harness; b) Julius-K9^®IDC harness; c) Julius-K9^®power harness; d) sketch of the Fressnapf own-branded harness. Note, that the Julius-K9^®harnesses were manufactured without the light-reflexive materials. Pictures a), b) and c) are reprinted from https://julius-k9.com/en/ under a CC BY license, with permission from Julius-K9^®, original copyright (1997–2020).

Measurement setup

The measurement system was an OptiTrack (NaturalPoint, Corvallis, Oregon, USA) three-dimensional optical motion capture system consisting of 18 Flex13 cameras and the Motive v1.10.3 software [20]. The software is responsible for the coordinated operation of the 18 cameras and for recording the 3-dimensional position of markers. This measuring system’s accuracy is sub-millimetre (3D MeanErr: 0.537±0.016 mm), and the measurements were carried out at a sampling frequency of 120 Hz [21]. The cameras cover a 4x2.5 m measuring area, the treadmill being placed in the middle of it. The measurements were carried out at the Department of Mechatronics, Optics and Engineering Informatics at Budapest University of Technology and Economics, Hungary.

Based on Hogy et al., twenty-five infra reflective markers were placed on the dog’s specific anatomical landmarks (Fig 3) [22]. The same marker-set arrangement was used for each measurement. On the thoracic limbs, markers were placed over the distal lateral aspect of the fifth metacarpal bone (FR5 and FL5), the ulnar styloid process (FR4 and FL4), the lateral epicondyle of the humerus (FR3 and FL3), the greater tubercle of the humerus (FR2 and FL2), and the dorsal aspect of the scapular spine (FR1 and FL1). On the pelvic limbs, markers were placed over the distal lateral aspect of the fifth metatarsal bone (BR5 and BL5), the lateral malleolus of the fibula (BR4 and BL4), the lateral femoral condyle (BR3 and BL3), the greater trochanter of the femur (BR2 and BL2), and the iliac crest (BR1 and BL1). On the spine, markers were placed over the sacral apex (S5), the dorsal spinous process of vertebra L7, the dorsal spinous process of vertebra T13, the dorsal spinous process of vertebra T1, and the occipital protuberance (S1). The markers were fixed on the dogs using adhesive tape. Each participating dog was short-haired, thus their fur did not influence marker placement or induced unnecessary marker movement. In cases where a harness covered the location of an anatomical landmark (FR2-FL2), the marker was placed on top of the harness, as close to the anatomical landmark as possible.

Post-processing

Exporting data

A technician first processed the recorded marker data in Motive as follows: markers were manually labelled according to the used marker-set (Fig 3) for each recording. Next, a section of homogeneous gait between receiving treats was selected for each trial, and exported into a text file containing metadata of the measurement in a header—like frame rate and total number of frames—and the marker position data for each frame. For all further calculations, MATLAB (R2020b) was used [23].

Filtering

Filtering marker data with a low-pass Butterworth filter (f_c = 5–6 Hz) is a long-accepted norm in human gait analysis [24–26]. However, when inspecting positional marker data before and after applying this filter, it is apparent that it is not suitable for canine measurements, especially for smaller dogs. Larger movements (typically in the cranial-caudal direction) have good signal-to-noise ratio to begin with, and thus are not significantly affected by the filtering (Fig 4 left). However, the smaller components (vertical direction) have useful data beyond 5 Hz, and their signal is significantly distorted, especially on smaller dogs (Fig 4 right).

Fig 4 — Ideal low-pass and 6th order Butterworth filters are shown in figure. Left: forward direction; Right: vertical direction.

A 6th order zero-lag (using the filtfilt method for eliminating phase delay [23]) Butterworth filter and an ‘ideal’ low-pass filter were compared with multiple cutoff frequencies (f_c). The ‘ideal’ low-pass filter was achieved by applying fast Fourier transform (FFT) to the data, removing the components on frequencies above f_c, and converting the resultant spectrum back to a time series with inverse FFT. Based on observing multiple cases, it was found that resulting information loss affected the smallest dog’s marker positions the most. Accordingly, f_c was tuned to provide good results for the smallest dog, assuming that noise is independent of the dog measured, and occupies the same frequencies for all trials. After examining multiple f_c values, 20 Hz was determined to be appropriate, where no relevant information seemed to be lost due to filtering. A forward and vertical landmark position component before and after filtering with this cutoff frequency can be seen on Fig 5.

Fig 5 — Ideal low-pass and 6th order zero-lag Butterworth filters are shown in figure. Left: forward direction; Right: vertical direction.

Both for f_c selection and the actual processing, non-harmonic components are removed before applying the filters by fitting a trend line to the marker position component and subtracting it. Afterwards, the trend is re-added to the position data. We found that the Butterworth filter could more smoothly follow the original signal, whereas the ‘ideal’ filter had trouble matching the signal’s endpoints, introducing artefacts at the beginning and end. Thus, the selected filter was a 6th order Butterworth filter at f_c = 20 Hz (Fig 5, green lines), which was applied for every maker position component.

Gait cycle segmentation

In the next step, the recording was segmented into complete gait cycles from which the spatio-temporal parameters can be calculated. Heel strike and toe-off events for all limbs have to be determined. Similarly to human gait, heel strike occurs when the foot is farthest forward compared to the hip/shoulder [27], while toe-off occurs while the foot is the furthest behind the given joint. The corresponding frame numbers are found by calculating the feet-shoulder/hip joint distances and finding the peaks with the findpeaks function [23]. The back right limb’s heel strike (based on marker BR5, see Fig 3) was chosen as the separating points of the gait cycles.

Measured and calculated parameters

Paw parameters

A common parameter to represent the change in the cyclic forward movement is the path traced by selected points, such as the head or the centre of mass, but any measured point can be drawn. Paw movement (plots of each limb’s fifth marker in the sagittal plane) was chosen as a way to represent the cyclic forward movement of the animal. Other common points used to represent the forward movement is the centre of mass or the head, but centre of mass is difficult to calculate with the given marker set, and head movements were found to be very inconsistent due to the measurement setup, as the head position was not ‘fixed’ compared to the body (the dogs were moving their head around as to better observe things).

Joint angle calculation

The 2D projections of joint angles are calculated for each data frame. Spine angles in the horizontal plane at T1, T13 and L7 vertebrae are defined on Fig 6a. Joint angles in the sagittal plane include the spinal angles at the T1, T13 and L7 vertebrae, and the upper (shoulder/hip), middle (elbow/stifle) and lower (left and right carpal/hock) joint angles for all four limbs (Fig 6b). Joint angles are calculated by taking the two coordinates of the three defining anatomical landmarks in the given plane for each data frame and calculating the angle between the connecting vectors. Detailed descriptions of the joint angles are given in Table 2.

Table 2. Description of the calculated joint angles.

Joint angle name	Description	Plane
T1 spinal angle (horizontal aspect)	angle of T1-S1 and T1-T13	dorsal/ horizontal
T13 spinal angle (horizontal aspect)	angle of T13-T1 and T13-L7	dorsal/ horizontal
L7 spinal angle (horizontal aspect)	angle of L7-T13 and L7-S5	dorsal/ horizontal
T1 spinal angle (sagittal aspect)	angle of T1-S1 and T1-T13	sagittal
T13 spinal angle (sagittal aspect)	angle of T13-T1 and T13-L7	sagittal
L7 spinal angle (sagittal aspect)	angle of L7-T13 and L7-S	sagittal
Shoulder angle (upper front joint angle)	angles of FR2-FR1 and FR2-FR3 (right side); angles of FL2-FL1 and FL2-FL3 (left side)	sagittal
Hip angle (upper back joint angle)	angles of BR2-BR1 and BR2-BR3 (right side); angles of BL2-BL1 and BL2-BL3 (left side)	sagittal
Elbow angle (middle front joint angle)	angles of FR3-FR4 and FR3-FR2 (right side); angles of FL3-FL4 and FL3-FL2 (left side)	sagittal
Stifle angle (middle back joint angle)	angles of BR3-BR4 and BR3-BR2 (right side); angles of BL3-BL4 and BL3-BL2 (left side)	sagittal
Carpus angle (lower front joint angle)	angles of FR4-FR5 and FR4-FR3 (right side); angles of FL4-FL5 and FL4-FL3 (left side)	sagittal
Tarsus angle (lower back joint angle)	angles of BR4-BR5 and BR4-BR3 (right side); angles of BL4-BL5 and BL4-BL3 (left side)	sagittal

Open in a new tab

The calculated joint angle time series are segmented according to the back right foot’s heel strike (the ends of gait cycles) resulting in one time series for a given joint angle per gait cycle. These are then re-interpolated to 101 equally spaced points giving the joint angle value at each percent of the gait cycle, from 0 to 100%. Multiple time series for a given joint can be represented with an average joint angle and a 95% confidence band.

Spatio-temporal parameters

With knowing the heel strike and toe-off locations, as well as the set speed of the treadmill (noted from the treadmill’s display), the spatio-temporal parameters of each gait cycle can be calculated (Table 3). ROM for every gait cycle for each joint angle is also calculated and evaluated the same way as spatio-temporal parameters. In total, 18 joint angle curves (detailed in Fig 6 and Table 2), 4 plots of the paw movement in the sagittal plane and 53 scalar parameters (35 spatio-temporal parameters detailed in Fig 7 and Table 3 and 18 joint angle ROM parameters) can be given for each gait cycle.

Table 3. Spatio-temporal parameters.

Joint angle name	Description	Dim
Stride time / Cycle time	length of the gait cycle in time	s
Cadence	number of steps per minute calculated as 4 ⋅ (60/Stridetime)	$\frac{steps}{time}$
Speed	speed of the forward movement of the animal, calculated as the average of the back and front stride distance divided by stride time	m/s
Stride distance (back and front)	distance from one heel strike to the next	m
Walking base (back and front)	distance of the given limb’s heel strike form the opposite side limbs heel strike in the direction of movement	mm
Step distance (for all four limbs)	distance of the given limb’s heel strike form the opposite side limbs heel strike in the direction of movement	m
Step height (for all four limbs)	the range of the vertical component of the 5th marker on the given limb (how high the dog lifts it’s paw)	mm
Swing time (for all four limbs)	the time when a given limb is not in contact with the ground (time from toe-off to heel strike)	s
Swing ratio (for all four limbs)	the ratio of the swing time compared to the full cycle time	- (%)
Stance time (for all four limbs)	the time when a given limb is in contact with the ground time from heel strike to toe-off)	s
Stance ratio (for all four limbs)	the ratio of the stance time compared to the full cycle time	- (%)
Paw travel distance (for all four limbs)	distance that the foot travels during one stride on the treadmill	m

Open in a new tab

Evaluation

Per-case report

The goal of the presented processing method is to allow for a thorough analysis and comparison of gaits. To help this, automatically generated reports were created that allows for the visual examinations of all parameters. The joint angles and the shape of the path traced by the paws during movement are represented as plots, while all of the spatio-temporal and ROM parameters are given in violin-like charts. Multiple trials can be displayed on the same plot, providing a basis for evaluating how a specific circumstance (free walk, free walk wearing the harness and leashed walk wearing the harness) affects a particular dog’s motion. An example of a report like this is given can be seen in Fig 8.

These reports provide useful insights into individual cases, e.g., selecting the best harness for a working dog. The used method could also be applied to other canine gait analysis cases e.g. comparing movement before and after treatment for lameness.

Statistical analysis

Individual reports do not provide insight into the overall effects of harnesses, which is the test case of this pilot study. For this purpose, statistical evaluations were used. First, we tested for normality of the spatio-temporal and ROM parameter distributions across gait cycles in the different measurement cases (the free, harness wearing and leashed scenarios) with the Anderson-Darling test (H0: The distribution is from a normal distribution, α = 0.05). Parameters that do not pass the normality test more often can indicate problem areas for the specific dog or case. Next, each parameter’s distributions for each dog-harness combination for the cases of reference vs harness, harness vs harness+leash, and reference vs harness+leash were tested for identical distributions. One test involves comparing the samples of one parameter collected over multiple gait cycles of the same dog in the two different trials. For this, the two-sample Kolmogorov-Smirnov test for identical distributions (H0: The two sample sets are from the same distribution, α = 0.05) were used. This test can show if either the expected value or the parameter’s variation has changed (it does not tell which one, however) and does not require the datasets to be normally distributed. In total 3 times 53 tests were needed per dog-harness combination.

Additionally, for all harness and harness+leash cases, root mean square (RMS) errors between the average joint angles were calculated as:

\begin{matrix} {RMS}_{t r i a l} = \sqrt{\frac{\sum_{i = 0}^{100} {(φ_{r e f, i} - φ_{t r i a l, i})}^{2}}{101}}, \end{matrix}

(1)

where trial represents a specific case (e.g. Dog 2 with the Julius-K9^®power harness and leash), φ_ref,i is the average joint angle of the reference trials of the given dog at the i-th integer percent of the gait cycle, and φ_trial,i is the average joint angle of the selected trial of the given dog at the i-th integer percent of the gait cycle. This value assigns a numerical score for how close a joint angle is to the reference.

Results

Tabular results of the 53 calculated parameters for all cases can be found in S1 Table. The table of results for the Anderson-Darling normality test can be found in S2 Table. Fig 9 shows for each parameter the percentage of cases where the parameter passed the Anderson Darling normality test (p > 0.05). Meanwhile, Fig 10 shows the inverse relation: how many parameters from a given trial passed the normality test.

Fig 9 — Percentage of cases (indicated by the distance from the centre) shows that a given parameter (spatio-temporal or joint range of motion, as named on the edge of the circle) passed the Anderson-Darling normality test (p > 0.05) (28 cases in total).

Fig 10 — Percentage of parameters of a given trial that passed the Anderson-Darling normality test (p > 0.05)(53 parameters in total).

The table of results for the two-sided Kolmogorov-Smirnov test can be found in S3 Table. Fig 11 shows in what percentage of cases each parameter passed the two-sided Kolmogorov-Smirnov test for identical distributions (p > 0.05). Fig 12 shows the inverse: how many parameters from a given configuration passed the Kolmogorov-Smirnov test.

Fig 12 — Percentage of parameters of a given case passed the two-sided Kolmogorov Smirnov test for identical distribution (p > 0.05)(53 parameters in total).

Finally, the RMS errors between the average joint angles can be seen in Table 4. In two trials, a spine marker was not visible on the recordings, so the corresponding cells are left empty.

Table 4. RMS error values for average joint angle differences.

Case	T1	T13	L7	T1	T13	L7	FR	FL	FR	FL	FR	FL	BR	BL	BR	BL	BR	BL
Case	hor.	hor.	hor.	sag.	sag.	sag.	sho.	sho.	elb.	elb.	car.	car.	hip	hip	sti.	sti.	hock	hock
Dog 1 K9 power	1.08	1.88	0.83	5.94	0.68	0.90	5.38	6.47	3.13	5.62	1.95	1.72	0.50	0.70	0.99	1.31	0.56	0.96
Dog 1 K9 power (leash)	3.41	4.20	4.04	16.48	6.42	3.25	7.09	10.22	8.23	9.95	15.73	12.13	2.59	2.94	4.14	3.86	7.35	6.16
Dog 1 K9 IDC	3.68	2.72	1.34	8.20	1.58	0.55	8.16	21.85	2.65	21.96	3.17	4.57	1.30	1.78	2.24	2.88	2.09	2.59
Dog 1 K9 IDC (leash)	1.79	3.61	4.13	13.23	5.91	3.13	17.90	13.87	16.79	14.06	22.60	22.73	2.74	2.32	7.22	5.77	11.66	8.18
Dog 1 K9 Duo-Flex	1.26	1.44	1.14	3.19	2.60	0.39	37.22	38.37	21.70	22.86	6.76	5.19	1.43	1.24	1.91	1.87	2.78	2.33
Dog 1 K9 Duo-Flex (leash)	1.95	2.23	4.42	10.00	6.65	6.74	63.26	63.75	35.63	33.16	15.50	14.04	3.11	3.22	9.54	6.98	10.89	7.96
Dog 2 K9 power	2.00	7.63	5.56	23.56	6.83	1.07	10.09	3.73	9.63	13.52	7.04	4.38	3.35	3.11	1.29	1.60	1.31	1.13
Dog 2 K9 power (leash)	5.53	9.42		17.04	10.96		7.54	4.62	5.39	4.68	7.25	6.40	5.41	11.10	6.43	9.26	6.01	14.32
Dog 2 K9 IDC	2.41	3.64	2.97	9.31	0.72	0.55	15.57	5.08	17.18	8.94	5.67	6.02	1.14	1.57	0.97	1.26	1.81	2.34
Dog 2 K9 IDC (leash)	3.23	5.55	5.04	20.54	5.88	3.21	12.11	6.89	16.70	11.87	5.97	6.76	4.00	6.82	3.82	4.14	4.92	7.42
Dog 2 K9 Duo-Flex	2.12	2.80	1.74	11.86	1.90	0.58	15.47	8.84	15.60	13.24	8.97	10.90	3.12	3.24	2.28	2.01	3.88	3.80
Dog 2 K9 Duo-Flex (leash)	4.45	11.75	7.95	14.63	5.52	5.03	7.63	8.58	10.10	10.49	13.34	15.11	5.83	12.90	6.60	6.55	7.36	8.75
Dog 3 K9 power	1.76	5.64	2.26	1.67	4.86	0.99	5.80	11.58	8.73	7.63	10.04	6.70	0.88	0.78	1.68	2.51	1.99	2.67
Dog 3 K9 power (leash)	5.24	10.52		3.97	22.15		11.98	14.01	18.30	14.00	36.71	31.34	3.01	2.69	3.28	3.91	4.09	4.50
Dog 3 K9 IDC	2.59	6.71	5.54	3.32	0.91	0.89	6.41	6.10	6.82	6.68	9.43	7.99	1.00	0.72	1.15	1.45	1.04	1.47
Dog 3 K9 IDC (leash)	4.46	9.02	3.79	6.64	4.03	2.90	10.76	6.69	9.09	5.93	16.47	13.00	3.98	4.37	8.18	10.10	8.61	9.68
Dog 4 K9 power	3.72	1.96	1.55	7.19	1.50	0.89	10.11	17.18	18.66	24.19	15.26	21.46	2.12	1.47	3.05	3.30	4.42	3.64
Dog 4 K9 power (leash)	2.69	2.26	2.99	11.80	1.54	2.67	6.38	21.34	24.51	33.84	13.31	14.89	1.71	1.79	6.47	6.40	7.72	6.83
Dog 4 K9 IDC	2.85	2.32	0.76	3.83	1.66	1.00	9.77	7.85	6.84	5.07	21.23	12.31	2.98	2.27	3.85	3.72	5.04	4.21
Dog 4 K9 IDC (leash)	2.68	4.84	3.68	5.11	1.58	1.57	12.81	15.41	16.81	26.18	19.48	11.78	2.35	1.67	5.12	5.76	6.19	6.06
Dog 4 K9 Duo-Flex	2.33	1.21	1.02	3.96	1.63	1.09	46.08	57.44	23.27	34.83	12.80	13.26	1.09	0.87	1.64	1.82	1.60	1.74
Dog 4 K9 Duo-Flex (leash)	2.21	1.63	2.93	9.32	3.02	4.21	68.75	79.10	37.14	48.67	10.83	10.85	1.77	1.55	4.50	6.14	5.41	6.25
Dog 4 Fressnapf	11.32	2.16	3.71	11.88	1.44	5.79	15.65	4.63	6.83	3.42	5.85	14.36	5.69	11.31	7.40	6.21	11.00	16.15
Dog 4 Fressnapf (leash)	8.16	4.62	1.61	9.63	2.24	9.45	27.26	11.87	6.37	7.15	7.53	14.39	6.09	12.43	3.97	3.83	15.83	21.98

Open in a new tab

Values are in degrees. Colours from green through yellow to red indicate lowest to highest RMS. Missing values indicate joint angles that could not be calculated for that case.

Discussion

This pilot study’s goal was to establish a measurement method capable of quantifying canine gait in detail, which can determine the spatio-temporal parameters of all four limbs, the joint angles of the major joints and spinal angles from the spatial coordinate of selected anatomical landmarks, captured throughout multiple gait cycles. Based on the markerset adapted from Hogy et al. (2013) (Fig 3), the 3D motion of 25 anatomical landmarks can be recorded with a motion capture system [22]. After determining an appropriate filtering method (a zero-lag 6th order Butterworth filter achieved with MatLab’s filtfilt method [23], f_c = 20 Hz, with the non-harmonic components removed before and re-added after by fitting a trend line to the marker position components), the recording is segmented according to gait cycles by determining the heel strike and toe-off events based on the distance of the feet and the shoulder/hip joint in the direction of movement. In total, 18 joint angle curves (detailed in Fig 6 and Table 2), four plots of the paw movement in the sagittal plane and 53 scalar parameters (35 spatio-temporal parameters detailed in Fig 7 and Table 3 and 18 joint angle ROM parameters detailed in Fig 6 and Table 2) can be given for each gait cycle based on the recorded marker positions. An example report detailing these parameters can be seen in Fig 8. Based on Fig 9, most parameters pass the test for normality in 60+% of the cases (there were 28 trials in total). The parameters failing the test more regularly are the Swing time parameters. This is likely due to the fact that swing times have minimal values (usually between 0.2 and 0.3 seconds), where the variation appears to have discreet values determined by the camera’s frame rate (the Anderson-Darling test is more geared towards practically continuous distributions) (Fig 13a).

Fig 13 — a) discreet values for swing time; b) negative swing ratio in case of Dog 2; c) distribution of T1 sagittal spine angle range of motion with a notable tails towards larger values; d) large tail towards lower values on the cycle time parameter of Dog 4 on the reference trial (red points).

Front Right Swing Time, Swing Ratio, Stance Time and Stance Ratio all failed to pass the test in case of dog 2. Examining these parameters show a few cases where swing and stance time became negative values generating outliers in the data. The outliers in stride and swing times also cause outliers in the same ratio parameters (Fig 13b). This indicates irregularities on the dog’s gait where the front left heel-strike event of the amble (typically very shortly after the back right heel strike) happened too early in some cycles, before the heel strike of the back right limb. This indicates possible injury or other type of irregularity in the front right limb of Dog 2.

Another parameter that often does not follow normal distribution is the ROM of the sagittal aspect of the spinal angle at the T1 vertebrae. This is due to the dogs raising or lowering their heads for several gait cycles at a time as they are looking at their owner with the treats in front of them, generating larger ROM values than usual (Fig 13c).

As for the per case comparison, 60–80% of the parameters passed the normality test in most cases. The most notable case is the reference trial for dog 4, where less than 40% of the parameters passed, like the Cycle time, stride length, speed parameters. In this case, several gait cycles appeared to be much shorter in time and step lengths than the rest (Fig 13d), affecting several other parameters too, and creating long tails on one end of the distributions. No such behaviour was displayed by the same dog when in any harness so the reason for this is unclear.

The developed canine gait analysis methodology was tested in comparing the effects of different designs of harnesses on gait (which can be characterised by the 53 scalar and 18 joint angle curve and 4 paw movement plot parameters described previously) and to study whether certain designs are better than others. Based on Fig 11, apart from speed, all parameters were affected in most cases (speed being constrained by the treadmill in this case). As expected, when comparing the reference to the leashed cases, the parameters pass the two sample Kolmogorov-Smirnov test in even fewer cases. Fig 12 is a good demonstration that all of the tested harnesses alter a large portion of the calculated gait parameters for each dog tested, and as such, none of them proved to be a ‘perfect harness’. Despite this, a conclusion regarding each dog can be drawn as to which harness could be preferred. For the smallest dog (Dog 3), the Julius-K9^®power harness seems to be the better fit. For Dog 4, the fastest walking dog, Duo-flex seems to be the better option in leash-less cases, and the Fressnapf own-branded harness for walking with a leash. In the case of Dog 2, IDC is the better choice, while for Dog 1, the Julius-K9^®power harness. The results do not indicate any connection between the preferable harness and the breed, size or walking speed of the dogs, nor would it be realistic to draw these conclusions based on the sample size of this study.

Table 4 shows two things. Firstly, as expected the highest deviation from the natural (reference motion) is in the thoracic limbs’ angles. In the most extreme cases of Dog 1 and Dog 4 (Duo-Flex cases), the large RMS errors result from the whole joint angle curve getting offset from the original (Fig 14). This is due to some markers having to be placed on top of the harness instead of directly on the dog. The positions of these markers compared to the actual location of the anatomical landmark is highly uncertain. Leash force could result in additional displacement. It has been previously shown for human gait analysis that deviations in marker placement results in similar offset errors in joint angles, but the general characteristic of the curve is unchanged [28].

A second observation based on Table 4 is that RMS errors of harness+leash cases are larger than just the harness cases. This applies for all joint angles, not just the ones potentially effected by the offset error from putting the marker on the harness. Based on the observations, the RMS error of joint angles numerically support the two intuitively known facts: harnesses mostly affect thoraic limbs more, and the addition of a leash further deviates the motion from the natural one.

Comparing the results with the study of Lafuente et al., the results can be said to agree [13]. Fig 11 shows that ROM of the shoulder joint was significantly affected by the inclusion of a harness compared to the reference case in virtually all trials, with the harness and leash cases showing smaller number of significantly different cases compared to harness only trials. Table 4 shows that the non-restrictive style harness (Duo-flex) shows significantly larger change on the shoulder joint angles compared to the other harnesses for dogs 1 and 4, while showing about the same for dog 2 (dog 3 was not tested in the duo-flex harness). This also agrees with the findings of Lafuente et al. stating that the non-restrictive harnesses are actually more restrictive in respect to the shoulder angles. However, as stated before these cases are the ones markers might have had to been placed on top of the marker, causing high uncertainty. Comparing the results with the study of Peham et al. [16], Fig 11 also confirms that harnesses significantly affect the movement of the spine, although the effect seems to be more pronounced in the horizontal plane compared to the sagittal motions.

To test the possible limitations of this method, four dogs with varying body shapes and sizes were measured comparing a number of different harnesses. The studied dogs did not provide enough data for any meaningful conclusions on evaluating different harnesses. Combined with the large number of parameters calculated, the amount of data is far less than necessary to draw clear conclusions about how individual parameters are affected. The difference in breeds and body sizes of the dogs further hinder drawing these conclusions. However, it was shown that the method itself is usable regardless of the body size of the dog, or the type of the harness.

A limitation of the method is using projected 2D joint angles instead of anatomically defined 3D joint angles (flexion-extension, abduction-adduction, internal-external rotation) for the limbs. Proper calculation of these angles require information about the 3D rigid-body motion of each segment of the limbs, which in turn require at least 3 markers on each body segment. While this would be plausible for larger dogs, in case of small dogs the markers would have to be too close for each other for our current setup. The possibility of anatomically defined joint angles on all sizes of dogs could be further explored with a more specialised setup, where the cameras are placed in the immediate surroundings of the treadmill. 2D joint angles are also more comparable with single camera measurements, where joint angles are determined on a single video recording of the dog from the side.

An other issue with examining harnesses specifically, is that some harnesses might end up covering an anatomical landmark of the dog (the FR2 anatomical landmark in most cases). In this study, we chose to place the marker on top of the harness in these cases, as close to the the landmark below as possible. However, this adds an extra layer on top of the skin that can shift compared to the landmark, causing the results to be inaccurate, and affecting the calculation of the front shoulder and elbow joint angles.

We aim to continue this research in a larger MoCap studio, where we can test more dogs within several groups of body sizes performing walk and trot on the ground instead of a treadmill to observe more natural movements. The anomalistic parameter distribution presented on Fig 13 should also be further studied, particularly in the case of getting negative values for the swing time and ratio parameters.

Conclusion

The novelty of this study is the large number of computable gait parameters established for canine gait analysis: spatio-temporal parameters for all four limbs, the sagittal aspect of the major limb joint angles, and the sagittal and horizontal aspects of spinal angles, as well as the path of any measured anatomical landmark (paw movement in the sagittal plane was highlighted here). An appropriate filtering process for the marker coordinates in case of dogs were also established. These calculated gait parameters were examined for four different dogs, comparing cases of movement on a treadmill in a natural and harness-wearing (with and without leash) state. This pilot study demonstrates that 3D gait analysis presents an opportunity for examining an extensive array of parameters and providing in-depth analysis for multiple purposes, whether it would be for examining the movements of a single dog for veterinary treatment or training, or studying the effects of different harness designs, with a more extensive parameter set than before. The results demonstrate that the method is applicable to various dog sizes and harnesses.

The results of the Kolmogorov-Smirnov test indicate that the least similar gait pattern is obtained when comparing the reference to the leashed cases. However, the tendencies based on the results show that all the tested harness altered the gait pattern for each dog and such there would be hard to find a one-size-fits-all option from them. For example some harnesses might perform better on smaller size dogs, while a different harness might be better suited for larger ones, and that is not even considering the different use cases, i.e., a walk in the city compared with a hike in the woods, or work for a service dog. This is an empirically accepted standpoint, but there is no research supporting it as of yet to the authors’ knowledge. More specific findings regarding the “goodness” of the harnesses would requires measurements with more participating dogs, which would be one of the possible advances in this present research. The method presented here can be utilised to evaluate harnesses on a dog-by-dog cases, selecting the most suitable for the given activities. The method can also be re-purposed to any case, where multiple sets of canine gaits need to be compared.

Supporting information

S1 Table. Table of calculated scalar parameters.

Tabular results of the 53 calculated scalar parameters for all measurement scenarios.

(PDF)

Click here for additional data file.^{(252.7KB, pdf)}

S2 Table. Tabular results of the Anderson-Darling normality test.

Tabular results of the p-values of the Anderson-Darling normality test for the 53 calculated scalar parameters for all measurement scenarios.

(PDF)

Click here for additional data file.^{(236.9KB, pdf)}

S3 Table. Tabular results of the two-sided Kolmogorov-Smirnov test.

Tabular results of the p-values of the two-sided Kolmogorov-Smirnov test for the 53 calculated scalar parameters comparing reference and harness trials, harness and harness+leash trials and reference and harness+leash trials.

(PDF)

Click here for additional data file.^{(283.2KB, pdf)}

Acknowledgments

The authors would like to thank Dr. Otília Biksi for her contribution in overseeing the experiments.

Data Availability

All relevant data are within the paper and its Supporting information files.

Funding Statement

The research reported in this paper and carried out at BME has been supported by the Hungarian Scientific Research Fund (OTKA), grant number: K135042 (http://nyilvanos.otka-palyazat.hu/index.php?menuid=930&num=135042&lang=EN), and the National Research, Development and Innovation Fund (TKP2020 NC, No. BME-NCS) based on the charter of bolster issued by the National Research, Development and Innovation Office under the auspices of the Ministry for Innovation and Technology, Hungary. The three harnesses from Julius-K9^® were provided free of charge by the manufacturer. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References

1. Keegan KG, Wilson DJ, Wilson DA, Barnett CD, Smith B. Effects of balancing and shoeing of the forelimb feet on kinematic gait analysis in five horses with navicular disease. Journal of equine veterinary science, 1998. 18(8): 522–527. doi: 10.1016/S0737-0806(98)80073-7 [DOI] [Google Scholar]
2. Agostinho FS, Rahal SC, Miqueleto NSML, Verdugo MR, Inamassu LR, El-Warrak AO. Kinematic analysis of Labrador Retrievers and Rottweilers trotting on a treadmill. Veterinary and Comparative Orthopaedics and Traumatology, 2011. 24(3): 185–191. doi: 10.3415/VCOT-10-03-0039 [DOI] [PubMed] [Google Scholar]
3. McLaughlin RM. Kinetic and kinematic gait analysis in dogs. Veterinary Clinics of North America: Small Animal Practice, 2001. 31(1): 193–201. doi: 10.1016/S0195-5616(01)50045-5 [DOI] [PubMed] [Google Scholar]
4. Whittle MW. Clinical gait analysis: A review. Human movement science, 1996. 15(3): 369–387. doi: 10.1016/0167-9457(96)00006-1 [DOI] [Google Scholar]
5. Budsberg SC, Verstraete MC, Soutas-Little RW. Force plate analysis of the walking gait in healthy dogs. American journal of veterinary research, 1987. 48(6): 915–918. [PubMed] [Google Scholar]
6. Bennett RL, DeCamp CE, Flo GL, Hauptman JG, Stajich M. Kinematic gait analysis in dogs with hip dysplasia. American journal of veterinary research, 1996. 57(7): 966–971. [PubMed] [Google Scholar]
7. DeCamp CE, Soutas–Little RW, Hauptman J, Olivier B, Braden T, Walton A. Kinematic gait analysis of the trot in healthy greyhounds. American journal of veterinary research, 1993. 54(4): 627–634. [PubMed] [Google Scholar]
8. DeCamp CE Kinetic and kinematic gait analysis and the assessment of lameness in the dog. Veterinary Clinics: Small Animal Practice, 1997. 27(4): 825–840. [DOI] [PubMed] [Google Scholar]
9. Carr BJ, Dycus DL. Canine gait analysis. Recovery & Rehab, 2016. 6(2): 93–100. [Google Scholar]
10. Foss K, da Costa RC, Moore S. Three–dimensional kinematic gait analysis of Doberman Pinschers with and without cervical spondylomyelopathy. Journal of veterinary internal medicine, 2013. 27(1): 112–119. doi: 10.1111/jvim.12012 [DOI] [PMC free article] [PubMed] [Google Scholar]
11. Garcia TC, Sturges BK, Stover SM, Aoki K, Liang JM, Reinhardt KB, Kapatkin AS. Forelimb brachial muscle activation patterns using surface electromyography and their relationship to kinematics in normal dogs walking and trotting. Comparative Exercise Physiology, 2014. 10(1): 13–22. doi: 10.3920/CEP13026 [DOI] [Google Scholar]
12. Surer E, Cereatti A, Evangelisti MA, Paolini G, Della Croce U, Manunta ML. A Canine Gait Analysis Protocol for Back Movement Assessment in German Shepherd Dogs. Veterinary sciences, 2020. Feb 19. doi: 10.3390/vetsci7010026 [DOI] [PMC free article] [PubMed] [Google Scholar]
13. Lafuente MP, Provis L, Schmalz EA. Effects of restrictive and non–restrictive harnesses on shoulder extension in dogs at walk and trot. Veterinary Record, 2019. 184(2): 64–64. doi: 10.1136/vr.104946 [DOI] [PubMed] [Google Scholar]
14.Carr BJ, Dresse K, Zink MC. The effects of five commercially available harnesses on canine gait. In: Proceedings of American College of Veterinary Surgeons 2016 Surgical Summit; 2016 Oct 6–8; Seattle, Washington. Veterinary Surgery; 2016.
15.Blake S, Williams R, De Godoy, RF. A Systematic Review of the Biomechanical Effects of Harness and Head-Collar use in Dogs. BioRxiv [Preprint]. 2019. bioRxiv 759258 [posted 2019 Sept 5; cited 2021 Jan 9]: [23 p.]. Available from: https://www.biorxiv.org/content/10.1101/759258v1.full.
16. Peham C, Limbeck S, Galla K, Bockstahler B. Kinematic analysis of the influence of three different guide dog harnesses on the movement of the spine. Wiener Tierarztliche Monatsschrift, 2013. 100(11): 306–312. [Google Scholar]
17. Peham C, Limbeck S, Galla K, Bockstahler B. Pressure distribution under three different types of harnesses used for guide dogs. The Veterinary Journal, 2013. 198 Suppl 1: e93–e98. doi: 10.1016/j.tvjl.2013.09.040 [DOI] [PubMed] [Google Scholar]
18. Foutz TL, Budsberg SC. Impact of wearing a service vest on three-dimensional truncal motion in dogs. American journal of veterinary research, 2020. 81(3): 210–219. doi: 10.2460/ajvr.81.3.210 [DOI] [PubMed] [Google Scholar]
19. Zink C, Carr BJ. Locomotion and Athletic Performance. In: Zink C, Van Dyke JB, editors. Canine Sports Medicine and Rehabilitation. Second Edition. New York: John Wiley & Sons, Inc; 2018. pp. 23–42. [Google Scholar]
20.Optitrack Motive [Internet] NaturalPoint, Inc. version 1.10.3 [cited 2021. Sept 29]. Available from: https://v110.wiki.optitrack.com/index.php.
21. Nagymáté G, Kiss RM. Motion capture system validation with surveying techniques. Materials Today: Proceedings, 2018. 5(13): 26501–26506. [Google Scholar]
22. Hogy SM, Worley DR, Jarvis SL, Hill AE, Reiser RF, Haussler KK. Kinematic and kinetic analysis of dogs during trotting after amputation of a pelvic limb. American journal of veterinary research, 2013. 74(9): 1164–1171. doi: 10.2460/ajvr.74.9.1164 [DOI] [PubMed] [Google Scholar]
23.MATLAB [Internet] Natick, Massachusetts: The MathWorks Inc. (R2020b) Update 2. version 9.9.0.1524771 [cited 2021. Jan 9]. Available from: https://mathworks.com/help/matlab/.
24. Holden JP, Orsini JA, Siegel KL, Kepple TM, Gerber LH, Stanhope SJ. Surface movement errors in shank kinematics and knee kinetics during gait. Gait & Posture, 1997. 5(3): 217–227. doi: 10.1016/S0966-6362(96)01088-0 [DOI] [Google Scholar]
25. King SL, Barton GJ, Ranganath LR. Interpreting sources of variation in clinical gait analysis: A case study. Gait & posture, 2017. 52: 1–4. doi: 10.1016/j.gaitpost.2016.10.022 [DOI] [PubMed] [Google Scholar]
26. Manal K, McClay I, Stanhope S, Richards J, Galinat B. Comparison of surface mounted markers and attachment methods in estimating tibial rotations during walking: an in vivo study. Gait & posture, 2000. 11(1): 38–45. doi: 10.1016/S0966-6362(99)00042-9 [DOI] [PubMed] [Google Scholar]
27. Zeni JA Jr, Richards JG, Higginson JS. Two simple methods for determining gait events during treadmill and overground walking using kinematic data. Gait & posture, 2008. 27(4): 710–714. doi: 10.1016/j.gaitpost.2007.07.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Rácz K, Nagymáté G, Kiss RM. Effect of anatomical landmark placement variation on the angular parameters of the lower extremities. In: 2017 13th IASTED International Conference on Biomedical Engineering (BioMed). IEEE, 2017 Feb. pp. 158–163.

PLoS One. doi: 10.1371/journal.pone.0264299.r001

Decision Letter 0

Ewa Tomaszewska

9 Sep 2021

PONE-D-21-02564Developing a detailed canine gait analysis method for evaluating harnesses: a pilot studyPLOS ONE

Dear Dr. Rita M. Kiss,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by October 15 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Ewa Tomaszewska, DVM Ph.D

Academic Editor

PLOS ONE

Additional Editor Comments (if provided):

Dear Authors,

In Reviewer`s opinion, the global quality of this study is ABOVE AVERAGE, but should be corrected in accordance to comments sent by Reviewer.

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf.

2. In your Methods section, please provide additional details regarding participant consent from the owners of the animals. In the ethics statement in the Methods and online submission information, please ensure that you have specified (1) whether consent was informed and (2) what type you obtained (for instance, written or verbal). If the need for consent was waived by the ethics committee, please include this information.

3. We note that Figures 1,2,3 and 4 in your submission contain copyrighted images. All PLOS content is published under the Creative Commons Attribution License (CC BY 4.0), which means that the manuscript, images, and Supporting Information files will be freely available online, and any third party is permitted to access, download, copy, distribute, and use these materials in any way, even commercially, with proper attribution. For more information, see our copyright guidelines: http://journals.plos.org/plosone/s/licenses-and-copyright.

We require you to either (1) present written permission from the copyright holder to publish these figures specifically under the CC BY 4.0 license, or (2) remove the figures from your submission:

a. You may seek permission from the original copyright holder of Figures 1,2,3 and 4 to publish the content specifically under the CC BY 4.0 license.

We recommend that you contact the original copyright holder with the Content Permission Form (http://journals.plos.org/plosone/s/file?id=7c09/content-permission-form.pdf) and the following text:

“I request permission for the open-access journal PLOS ONE to publish XXX under the Creative Commons Attribution License (CCAL) CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). Please be aware that this license allows unrestricted use and distribution, even commercially, by third parties. Please reply and provide explicit written permission to publish XXX under a CC BY license and complete the attached form.”

Please upload the completed Content Permission Form or other proof of granted permissions as an "Other" file with your submission.

In the figure caption of the copyrighted figure, please include the following text: “Reprinted from [ref] under a CC BY license, with permission from [name of publisher], original copyright [original copyright year].”

b. If you are unable to obtain permission from the original copyright holder to publish these figures under the CC BY 4.0 license or if the copyright holder’s requirements are incompatible with the CC BY 4.0 license, please either i) remove the figure or ii) supply a replacement figure that complies with the CC BY 4.0 license. Please check copyright information on all replacement figures and update the figure caption with source information. If applicable, please specify in the figure caption text when a figure is similar but not identical to the original image and is therefore for illustrative purposes only.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: This study sought to present a detailed method for canine gait acquisition and analysis addressing the need to correctly assess dog harnesses, which should adapt to the breed, size, and use scenarios. In particular, the authors hypothesized that the presence of the harnesses can alter the canine gait. For this reason, the author implemented this experimental pilot study to quantify how different harnesses affected the dogs’ kinematics during walking with respect to the unleashed condition. Four trained dogs and different harnesses (three provided by Juliuus-K9 and a one custom-made) with leash and without leash were included in the analysis. The study was approved by an Ethical Committee in 2015. The acquisition protocol was based on the use of 25 markers and allowed to estimate 18 joint angles, 4 paw movement paths in the sagittal plane and a wide set of spatio-temporal parameters, which can be normalized on multiple gait cycles. The authors reported information about marker trajectories processing and results are then reported with respect to the identified breeds and harnesses. The authors claimed that the methods they proposed can be used in comparative assessments; furthermore, they reported that all the analysed harnesses altered dogs kinematics during plain walking on treadmill, but the system was able to provide useful information about a possible optimal choice. All the data are available without restriction.

General Comment

The hypothesis at the basis of this paper is clearly reported as far as the main objective. Both the experimental phase and data analysis are written with a good level of details; the analysis in particular was performed very well, and the synthesis of the obtained results is extremely appreciable. Although the methodology in itself is not that innovative, the application to canine movements provides useful hints and novel perspective.

The structure of the article seems to be precise (Abstract, Introduction, Methodology [with subheadings], Results, Discussion).

Experimental phase and data analysis seem to be clearly reported and are coherent with the work objectives. Several minor concerns are hereinafter reported.

The use of the English language seems to be correct.

Specific Comments

Title

Ok. I would only consider to change “developing” with “development of”.

Abstract

In general, this section is quite ok. Please, could you report some quantitative information about the most interesting findings you obtained. It is important to get – even from the abstract – why you thought that your method is reliable and the differences you were able to highlight.

Introduction

In general also this section is ok; I would only state better the main limitations of the actual studies and the main novelty and innovation of the methodology you proposed, not only in light of the specific application you dealt with (i.e., harnesses).

Methods

• Page 3/15 Line 79-80: Please justify the number of dogs you involved in this study; although this is a pilot study, the reader needs further information about the choice of the canine subjects and how this can affect the possible generalization of the main findings (to be discuss in the Discussion section).

• Page 3 Line 78. A period “.” seems to miss at the end of the sentence.

• Page 3 Table 1. Please provide further information about the gait patterns.

• Page 4/15 Line 91: Please give further information about the third-party harness. Why did you call it “third-party”? Were Julius-K9s’ ones not third-party harnesses? In figure 1 there is no representation of this forth harness.

• Page 4/15 Line 101: I think that there is a typo here “[refence: Motive]”, and a correctly reported reference.

• Page 4/15 Line 103: Please provide numerical information about the accuracy you obtained in your specific setup after volume calibration.

• Page 4/15 Line 122-124: Please justify the choice of placing the marker on the harness here or discuss any possible issue in the Discussion section. Harness could significantly move with respect to the dog’s body, couldn’t it?

• Page 4/15 Line 127: Please provide more detail about the “clean up” phase.

• Page 5/15 Line 183: Please justify the use of 2D projections, or otherwise, discuss this as a possibile actual limitation of this study.

Results

Very well reported both graphically and in the main text.

Discussion

In general, this section is ok since it discuss your main findings. However is quite missing a comparison with the current literature, at least, where information are available. Furthermore, you tried to underline possible limitation at page 12/15 line 331-338, but you have to underline better any issue that can limit the possible generalization of your results (including the methodological ones, as – for instance – the choice of estimating 2D joint angles).

Conclusions

• Page 13/15 Line 359-362: This sentence can be hardly supported by the results of this only study. If there are differences, with such a small sample size it is not quite possible to reliably ascribe them to the only type of harnesses.

References

The references to previous works seem to be precise, wide and up-to-date.

Figures

Very good.

Tables

Good.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Nicola Francesco Lopomo

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

PLoS One. 2022 Mar 9;17(3):e0264299. doi: 10.1371/journal.pone.0264299.r002

Author response to Decision Letter 0

5 Oct 2021

Respose to Reviewer

Dear Dr. Lopomo,

We have received your review of your article, and we would like to thank you for your extraordinarily thorough and relevant questions and suggestions! In the following section we will address each point you raised to the best of our abilities. Please find the full response with all the listed modifications in the attached PDF!

Title:

I would only consider to change “developing” with “development of”.

We do agree with this suggestion, and modified the title accordingly!

Abstract:

Thank you for the insightful comment. We have extended the abstract with some more detail and qualitative results of the statistical analysis. Also, we‘ve reworded the beginning not to exceed 300 words.

Introduction:

Thank you for briging our attention tot his point! We have changed the wording of the last paragraph of the introduction, to better reflect the novelty of our study, which is the level of complexity in the analysed parameters, and to point out that this method can also be used in research of canine motion that does not deal with harnesses!

Methods:

1. Page 3/15 Line 79-80: Please justify the number of dogs you involved in this study; although this is a pilot study, the reader needs further information about the choice of the canine subjects and how this can affect the possible generalization of the main findings (to be discuss in the Discussion section).

We have no other justification for it other than that’s how many we could find whose owners were willing to participate, and train their dogs beforehand for walking on a treadmill. We changed the wording for this to be more clearer in the paper.

2. Page 3 Line 78. A period “.” seems to miss at the end of the sentence.

Thank you, we added the missing period!

3. Page 3 Table 1. Please provide further information about the gait patterns.

We have added a reference with detailed descriptions of every canine gait pattern.

(Table 1. Participating dogs and used harnesses. Gait patterns were identified by eye by an expert. Detailed description of each gait pattern can be found in [18].

[...]

18. Zink C, Carr BJ. Locomotion and Athletic Performance. In: Zink C, Van DykeJB, editors. Canine Sports Medicine and Rehabilitation, Second Edition. NewYork, USA: John Wiley & Sons, Inc; 2018. pp. 23–42.)

4.Page 4/15 Line 91: Please give further information about the third-party harness. Why did you call it “third-party”? Were Julius-K9s’ ones not third-party harnesses? In figure 1 there is no representation of this forth harness.

The owner of dog 4 brought this harness to the measurement, and it was a generic non-restrictive type. Since the owner had no information about the manufacturer, we were not able to name any. This also means that we do not have a right to publish any picture of it. One for sure, it was not a Julis-K9 manufactured harness. Thank you for bringing this shortcoming to our attention; we have extended the description in the Methods section.

5. Page 4/15 Line 101: I think that there is a typo here “[refence: Motive]”, and a correctly reported reference.

Thank you for this remark. Further information about the calibration accuracy has been added to this section.

6. Page 4/15 Line 122-124: Please justify the choice of placing the marker on the harness here or discuss any possible issue in the Discussion section. Harness could significantly move with respect to the dog’s body, couldn’t it?

We strongly agree that this point should have been brought up in the discussion as a limitation of the study from the beginning. The harness can indeed move compared to the body considerably, but unfortunately in cases where the harness covered an anatomical landmark, we had no other way to get approximate position data of the given anatomical landmark (FR2) with the given measurement setup.

7. Page 4/15 Line 127: Please provide more detail about the “clean up” phase.

Unfortunately our previous wording made it look like there was some additional clean up that took place before labelling and exporting relevant sections of data, when in reality the labelling and exporting process was the “clean up”. Thank you for pointing out the ambiguous wording of this subsection! We have rewritten it to more clearly describe the process.

(A technician first processed the recorded marker data in Motive as follows: markers were labelled according to the used marker-set (Fig. 3) for each recording. Next, a section of homogeneous gait between receiving treats was selected for each trial, and exported into a text file containing metadata of the measurement in a header – like frame rate and130total number of frames – and the marker position data for each frame. For all further calculations, MATLAB (R2020b) was used [16].)

8. Page 5/15 Line 183: Please justify the use of 2D projections, or otherwise, discuss this as a possibile actual limitation of this study.

While we would have liked to used proper anatomical joint angles, the marker configuration required for it would have been much more complex, and very impractical on the smaller dogs with the camera system of our laboratory. We have added a paragraph to the discussion about this limitation explaining the requirements for calculating 3D join angles!

(A limitation of the method is using projected 2D joint angles instead of anatomically defined 3D joint angles (flexion-extension, abduction-adduction, internal-external rotation) for the limbs. Proper calculation of these angles require information about the 3D rigid-body motion of each segment of the limbs, which in turn require at least 3 markers on each body segment. While this would be plausible for larger dogs, in case of small dogs the markers would have to be too close for each other for our current setup. The possibility of anatomically defined joint angles on all sizes of dogs could be further explored with a more specialised setup, where the cameras are placed in the immediate surroundings of the treadmill. 2D joint angles are also more comparable with single camera measurements, where joint angles are determined on a single video recording of the dog from the side.)

Discussion:

Thank you for pointing out the missing comparison with the literature! Although not a lot of data can be found on the effects of harnesses, those we found do not contradict our results. A section has been added to the discussion comparing the results.

The limitations noted in previous suggestions (2D projections of joint angles and marker placement on harness) has also been addressed in the discussion.

Conclusion:

Page 13/15 Line 359-362: This sentence can be hardly supported by the results of this only study. If there are differences, with such a small sample size it is not quite possible to reliably ascribe them to the only type of harnesses.

Thank you for this observation. We have changed the wording of the last paragraph of the Conclusion, and the statement improved, as can be supported by our results accordingly. Some possible advances of the study have also been added to the paragraph.

Based on your comments and suggestions we have revised our manuscript, and we hope that our revisions prove to be satisfactory! Thank you again for your insightful critique. It helped us greatly in bringing our research up to par with the standards expected by the scientific community!

Best regards:

Zsófia Pálya, Kristóf Rácz, Gergely Nagymáté & Rita M. Kiss

Attachment

Submitted filename: rev2 - response_for_editor_K9.pdf

Click here for additional data file.^{(226.1KB, pdf)}

PLoS One. doi: 10.1371/journal.pone.0264299.r003

Decision Letter 1

Ewa Tomaszewska

27 Oct 2021

PONE-D-21-02564R1Development of a detailed canine gait analysis method for evaluating harnesses: a pilot studyPLOS ONE

Dear Dr. Rita M. Kiss,

Please submit your revised manuscript by November 15 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

We look forward to receiving your revised manuscript.

Kind regards,

Ewa Tomaszewska, DVM Ph.D

Academic Editor

PLOS ONE

Journal Requirements:

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

Reviewer #1: Partly

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

Reviewer #1: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

Reviewer #1: Yes

**********

6. Review Comments to the Author

Reviewer #1: I would like to thank the authors for the great effort they realized to answer all the concerns arisen during the first round of review.

However, several issues are still open and need further clarification:

Question: 1. Page 3/15 Line 79-80: Please justify the number of dogs you involved in this study; although this is a pilot study, the reader needs further information about the choice of the canine subjects and how this can affect the possible generalization of the main findings (to be discuss in the Discussion section).

Answer: We have no other justification for it other than that’s how many we could find whose owners were willing to participate, and train their dogs beforehand for walking on a treadmill. We changed the wording for this to be more clearer in the paper.

Further Comment: I guess that the choice of the number and breeds of the dogs should be justified better; a scientific approach requires that the sample size and characteristics must be defined before recruitment starts. Since the number of dog is extremely reduced and this could impact the possibility to generalize your approach, please state at your best the hypotheses at the basis of your choice.

Question: 4.Page 4/15 Line 91: Please give further information about the third-party harness. Why did you call it “third-party”? Were Julius-K9s’ ones not third-party harnesses? In figure 1 there is no representation of this forth harness.

Answer: The owner of dog 4 brought this harness to the measurement, and it was a generic non-restrictive type. Since the owner had no information about the manufacturer, we were not able to name any. This also means that we do not have a right to publish any picture of it. One for sure, it was not a Julis-K9 manufactured harness. Thank you for bringing this shortcoming to our attention; we have extended the description in the Methods section.

Further Comment: Unfortunately you did not get my hint. I underlined the fact that your study, although you reported that it is not, seems to be sponsored by Julius-K9 (as you reported, however, the three harnesses from Julius-K9 was provided free of charge by the manufacturer...that is a sort of sponsorship). For sure, you can, but you have to report better the main hypotheses at the basis of your research and why you did chose only Julus-K9 harnesses and, more specifically, those models. Since this is not a sponsored study, the 3rd party harness should be treated as those ones provided by Julius-K9, maybe using it as reference. You can provide a wider description and a sketch of this harness, if useful to better understand your results.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Nicola Francesco Lopomo

PLoS One. 2022 Mar 9;17(3):e0264299. doi: 10.1371/journal.pone.0264299.r004

Author response to Decision Letter 1

4 Jan 2022

Revision Round 2:

Dear Dr. Lopomo,

We have studied your suggestions and we would like to thank you for your constructive comments again. The responses of all the comments and recommendations are listed below.

Comment #1:

Question: Page 3/15 Line 79-80: Please justify the number of dogs you involved in this study; although this is a pilot study, the reader needs further information about the choice of the canine subjects and how this can affect the possible generalization of the main findings (to be discuss in the Discussion section).

We do understand the concern about the sample size of the pilot study. We did aim to have at least one small, medium and large size dog in the study to confirm the method is suitable for all sizes. Thankfully, all categories were represented with the dogs we could recruit. We added a few regards about this to the text of the manuscript.

Comment #2:

Thank you for this insightfull remark. We added some more general regards about the selected harnesses. Moreover, we would like to point out that in the case of Julius-K9 company the production takes place in Hungary. Upon our request they offered to manufacture the examined harrnesses without the reflective elements, thus facilitating our research work. In connection with harness bring by Dog 4 owner, we were also possible to find out where it was purchased and whether it was an own-branded product. In light of this, the “3rd party” name has been changed to “Fressnapf own-branded”. Based on the measurement photos and the suggestion, a sketch of this harness was added to the Figure 1.

Based on your comments and suggestions we have revised our manuscript, and we hope that our revisions proves to be satisfactory! Thank you again for your insightful critique. It helped us greatly in bringing our research up to par with the standards expected by the scientific community!

Best regards:

Zsófia Pálya, Kristóf Rácz, Gergely Nagymáté & Rita M. Kiss

Attachment

Submitted filename: rev3 - response_for_editor_K9.pdf

Click here for additional data file.^{(175.7KB, pdf)}

PLoS One. doi: 10.1371/journal.pone.0264299.r005

Decision Letter 2

Ewa Tomaszewska

9 Feb 2022

Development of a detailed canine gait analysis method for evaluating harnesses: a pilot study

PONE-D-21-02564R2

Dear Dr. Rita M. Kiss,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Ewa Tomaszewska, DVM Ph.D

Academic Editor

PLOS ONE

PLoS One. doi: 10.1371/journal.pone.0264299.r006

Acceptance letter

Ewa Tomaszewska

14 Feb 2022

PONE-D-21-02564R2

Development of a detailed canine gait analysis method for evaluating harnesses: a pilot study

Dear Dr. Kiss:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Professor Ewa Tomaszewska

Academic Editor

PLOS ONE

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

S1 Table. Table of calculated scalar parameters.

Tabular results of the 53 calculated scalar parameters for all measurement scenarios.

(PDF)

Click here for additional data file.^{(252.7KB, pdf)}

S2 Table. Tabular results of the Anderson-Darling normality test.

Tabular results of the p-values of the Anderson-Darling normality test for the 53 calculated scalar parameters for all measurement scenarios.

(PDF)

Click here for additional data file.^{(236.9KB, pdf)}

S3 Table. Tabular results of the two-sided Kolmogorov-Smirnov test.

(PDF)

Click here for additional data file.^{(283.2KB, pdf)}

Attachment

Submitted filename: rev2 - response_for_editor_K9.pdf

Click here for additional data file.^{(226.1KB, pdf)}

Attachment

Submitted filename: rev3 - response_for_editor_K9.pdf

Click here for additional data file.^{(175.7KB, pdf)}

Data Availability Statement

All relevant data are within the paper and its Supporting information files.

[pone.0264299.ref001] 1. Keegan KG, Wilson DJ, Wilson DA, Barnett CD, Smith B. Effects of balancing and shoeing of the forelimb feet on kinematic gait analysis in five horses with navicular disease. Journal of equine veterinary science, 1998. 18(8): 522–527. doi: 10.1016/S0737-0806(98)80073-7 [DOI] [Google Scholar]

[pone.0264299.ref002] 2. Agostinho FS, Rahal SC, Miqueleto NSML, Verdugo MR, Inamassu LR, El-Warrak AO. Kinematic analysis of Labrador Retrievers and Rottweilers trotting on a treadmill. Veterinary and Comparative Orthopaedics and Traumatology, 2011. 24(3): 185–191. doi: 10.3415/VCOT-10-03-0039 [DOI] [PubMed] [Google Scholar]

[pone.0264299.ref003] 3. McLaughlin RM. Kinetic and kinematic gait analysis in dogs. Veterinary Clinics of North America: Small Animal Practice, 2001. 31(1): 193–201. doi: 10.1016/S0195-5616(01)50045-5 [DOI] [PubMed] [Google Scholar]

[pone.0264299.ref004] 4. Whittle MW. Clinical gait analysis: A review. Human movement science, 1996. 15(3): 369–387. doi: 10.1016/0167-9457(96)00006-1 [DOI] [Google Scholar]

[pone.0264299.ref005] 5. Budsberg SC, Verstraete MC, Soutas-Little RW. Force plate analysis of the walking gait in healthy dogs. American journal of veterinary research, 1987. 48(6): 915–918. [PubMed] [Google Scholar]

[pone.0264299.ref006] 6. Bennett RL, DeCamp CE, Flo GL, Hauptman JG, Stajich M. Kinematic gait analysis in dogs with hip dysplasia. American journal of veterinary research, 1996. 57(7): 966–971. [PubMed] [Google Scholar]

[pone.0264299.ref007] 7. DeCamp CE, Soutas–Little RW, Hauptman J, Olivier B, Braden T, Walton A. Kinematic gait analysis of the trot in healthy greyhounds. American journal of veterinary research, 1993. 54(4): 627–634. [PubMed] [Google Scholar]

[pone.0264299.ref008] 8. DeCamp CE Kinetic and kinematic gait analysis and the assessment of lameness in the dog. Veterinary Clinics: Small Animal Practice, 1997. 27(4): 825–840. [DOI] [PubMed] [Google Scholar]

[pone.0264299.ref009] 9. Carr BJ, Dycus DL. Canine gait analysis. Recovery & Rehab, 2016. 6(2): 93–100. [Google Scholar]

[pone.0264299.ref010] 10. Foss K, da Costa RC, Moore S. Three–dimensional kinematic gait analysis of Doberman Pinschers with and without cervical spondylomyelopathy. Journal of veterinary internal medicine, 2013. 27(1): 112–119. doi: 10.1111/jvim.12012 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0264299.ref011] 11. Garcia TC, Sturges BK, Stover SM, Aoki K, Liang JM, Reinhardt KB, Kapatkin AS. Forelimb brachial muscle activation patterns using surface electromyography and their relationship to kinematics in normal dogs walking and trotting. Comparative Exercise Physiology, 2014. 10(1): 13–22. doi: 10.3920/CEP13026 [DOI] [Google Scholar]

[pone.0264299.ref012] 12. Surer E, Cereatti A, Evangelisti MA, Paolini G, Della Croce U, Manunta ML. A Canine Gait Analysis Protocol for Back Movement Assessment in German Shepherd Dogs. Veterinary sciences, 2020. Feb 19. doi: 10.3390/vetsci7010026 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0264299.ref013] 13. Lafuente MP, Provis L, Schmalz EA. Effects of restrictive and non–restrictive harnesses on shoulder extension in dogs at walk and trot. Veterinary Record, 2019. 184(2): 64–64. doi: 10.1136/vr.104946 [DOI] [PubMed] [Google Scholar]

[pone.0264299.ref014] 14.Carr BJ, Dresse K, Zink MC. The effects of five commercially available harnesses on canine gait. In: Proceedings of American College of Veterinary Surgeons 2016 Surgical Summit; 2016 Oct 6–8; Seattle, Washington. Veterinary Surgery; 2016.

[pone.0264299.ref015] 15.Blake S, Williams R, De Godoy, RF. A Systematic Review of the Biomechanical Effects of Harness and Head-Collar use in Dogs. BioRxiv [Preprint]. 2019. bioRxiv 759258 [posted 2019 Sept 5; cited 2021 Jan 9]: [23 p.]. Available from: https://www.biorxiv.org/content/10.1101/759258v1.full.

[pone.0264299.ref016] 16. Peham C, Limbeck S, Galla K, Bockstahler B. Kinematic analysis of the influence of three different guide dog harnesses on the movement of the spine. Wiener Tierarztliche Monatsschrift, 2013. 100(11): 306–312. [Google Scholar]

[pone.0264299.ref017] 17. Peham C, Limbeck S, Galla K, Bockstahler B. Pressure distribution under three different types of harnesses used for guide dogs. The Veterinary Journal, 2013. 198 Suppl 1: e93–e98. doi: 10.1016/j.tvjl.2013.09.040 [DOI] [PubMed] [Google Scholar]

[pone.0264299.ref018] 18. Foutz TL, Budsberg SC. Impact of wearing a service vest on three-dimensional truncal motion in dogs. American journal of veterinary research, 2020. 81(3): 210–219. doi: 10.2460/ajvr.81.3.210 [DOI] [PubMed] [Google Scholar]

[pone.0264299.ref019] 19. Zink C, Carr BJ. Locomotion and Athletic Performance. In: Zink C, Van Dyke JB, editors. Canine Sports Medicine and Rehabilitation. Second Edition. New York: John Wiley & Sons, Inc; 2018. pp. 23–42. [Google Scholar]

[pone.0264299.ref020] 20.Optitrack Motive [Internet] NaturalPoint, Inc. version 1.10.3 [cited 2021. Sept 29]. Available from: https://v110.wiki.optitrack.com/index.php.

[pone.0264299.ref021] 21. Nagymáté G, Kiss RM. Motion capture system validation with surveying techniques. Materials Today: Proceedings, 2018. 5(13): 26501–26506. [Google Scholar]

[pone.0264299.ref022] 22. Hogy SM, Worley DR, Jarvis SL, Hill AE, Reiser RF, Haussler KK. Kinematic and kinetic analysis of dogs during trotting after amputation of a pelvic limb. American journal of veterinary research, 2013. 74(9): 1164–1171. doi: 10.2460/ajvr.74.9.1164 [DOI] [PubMed] [Google Scholar]

[pone.0264299.ref023] 23.MATLAB [Internet] Natick, Massachusetts: The MathWorks Inc. (R2020b) Update 2. version 9.9.0.1524771 [cited 2021. Jan 9]. Available from: https://mathworks.com/help/matlab/.

[pone.0264299.ref024] 24. Holden JP, Orsini JA, Siegel KL, Kepple TM, Gerber LH, Stanhope SJ. Surface movement errors in shank kinematics and knee kinetics during gait. Gait & Posture, 1997. 5(3): 217–227. doi: 10.1016/S0966-6362(96)01088-0 [DOI] [Google Scholar]

[pone.0264299.ref025] 25. King SL, Barton GJ, Ranganath LR. Interpreting sources of variation in clinical gait analysis: A case study. Gait & posture, 2017. 52: 1–4. doi: 10.1016/j.gaitpost.2016.10.022 [DOI] [PubMed] [Google Scholar]

[pone.0264299.ref026] 26. Manal K, McClay I, Stanhope S, Richards J, Galinat B. Comparison of surface mounted markers and attachment methods in estimating tibial rotations during walking: an in vivo study. Gait & posture, 2000. 11(1): 38–45. doi: 10.1016/S0966-6362(99)00042-9 [DOI] [PubMed] [Google Scholar]

[pone.0264299.ref027] 27. Zeni JA Jr, Richards JG, Higginson JS. Two simple methods for determining gait events during treadmill and overground walking using kinematic data. Gait & posture, 2008. 27(4): 710–714. doi: 10.1016/j.gaitpost.2007.07.007 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0264299.ref028] 28.Rácz K, Nagymáté G, Kiss RM. Effect of anatomical landmark placement variation on the angular parameters of the lower extremities. In: 2017 13th IASTED International Conference on Biomedical Engineering (BioMed). IEEE, 2017 Feb. pp. 158–163.

PERMALINK

Development of a detailed canine gait analysis method for evaluating harnesses: A pilot study

Zsófia Pálya

Kristóf Rácz

Gergely Nagymáté

Rita M Kiss

Roles

Abstract

Introduction

Methods

Study design

Fig 1. Measurement arrangement.

Table 1. Participating dogs and used harnesses.

Dog studied

Harnesses studied

Fig 2. Studied Julius-K9® harnesses.

Measurement setup

Fig 3. Marker-set of 25 reflexive markers.

Post-processing

Exporting data

Filtering

Fig 4. Position data of FR5 marker of the smallest dog filtered with fc = 5 Hz.

Fig 5. Position data of FR5 marker of the smallest dog filtered with fc = 20 Hz.

Gait cycle segmentation

Measured and calculated parameters

Paw parameters

Joint angle calculation

Fig 6. Joint angles in the horizontal and sagittal plane.

Table 2. Description of the calculated joint angles.

Spatio-temporal parameters

Table 3. Spatio-temporal parameters.

Fig 7. Distance type parameters in the horizontal plane.

Evaluation

Per-case report

Fig 8. An example of a report comparing free movement with harness and harness + leash scenarios.

Statistical analysis

Results

Fig 9. Result of Anderson-Darling normality test for each parameter.

Fig 10. Result of Anderson-Darling normality test for each trial.

Fig 11. Result of Kolmogorov-Smirnov test for each parameter grouped according to the different configurations.

Fig 12. Result of Kolmogorov-Smirnov test for each trial grouped according to the different configurations.

Table 4. RMS error values for average joint angle differences.

Discussion

Fig 13. Anomalistic distributions.

Fig 14. Offset in the shoulder angles in the case of dog 4 Duo-Flex.

Conclusion

Supporting information

Acknowledgments

Data Availability

Funding Statement

References

Decision Letter 0

Ewa Tomaszewska

Roles

Author response to Decision Letter 0

Decision Letter 1

Ewa Tomaszewska

Roles

Author response to Decision Letter 1

Decision Letter 2

Ewa Tomaszewska

Roles

Acceptance letter

Ewa Tomaszewska

Roles

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases

Fig 2. Studied Julius-K9^® harnesses.

Fig 4. Position data of FR5 marker of the smallest dog filtered with f_c = 5 Hz.

Fig 5. Position data of FR5 marker of the smallest dog filtered with f_c = 20 Hz.