Evaluation of a collar-mounted accelerometer for detecting seizure activity in dogs

The Use of Triaxial Accelerometers and Machine Learning Algorithms for Behavioural Identification in Domestic Dogs (Canis familiaris): A Validation Study

Assessing the behaviour and physical attributes of domesticated dogs is critical for predicting the suitability of animals for companionship or specific roles such as hunting, military or service. Common methods of behavioural assessment can be time consuming, labour-intensive, and subject to bias, making large-scale and rapid implementation challenging. Objective, practical and time effective behaviour measures may be facilitated by remote and automated devices such as accelerometers. This study, therefore, aimed to validate the ActiGraph® accelerometer as a tool for behavioural classification. This study used a machine learning method that identified nine dog behaviours with an overall accuracy of 74% (range for each behaviour was 54 to 93%). In addition, overall body dynamic acceleration was found to be correlated with the amount of time spent exhibiting active behaviours (barking, locomotion, scratching, sniffing, and standing; R2 = 0.91, p < 0.001). Machine learning was an effective method to build a model to classify behaviours such as barking, defecating, drinking, eating, locomotion, resting-asleep, resting-alert, sniffing, and standing with high overall accuracy whilst maintaining a large behavioural repertoire.

Decoding the Gut Microbiome in Companion Animals: Impacts and Innovations

The changing notion of "companion animals" and their increasing global status as family members underscores the dynamic interaction between gut microbiota and host health. This review provides a comprehensive understanding of the intricate microbial ecology within companion animals required to maintain overall health and prevent disease. Exploration of specific diseases and syndromes linked to gut microbiome alterations (dysbiosis), such as inflammatory bowel disease, obesity, and neurological conditions like epilepsy, are highlighted. In addition, this review provides an analysis of the various factors that impact the abundance of the gut microbiome like age, breed, habitual diet, and microbe-targeted interventions, such as probiotics. Detection methods including PCR-based algorithms, fluorescence in situ hybridisation, and 16S rRNA gene sequencing are reviewed, along with their limitations and the need for future advancements. Prospects for longitudinal investigations, functional dynamics exploration, and accurate identification of microbial signatures associated with specific health problems offer promising directions for future research. In summary, it is an attempt to provide a deeper insight into the orchestration of multiple microbial species shaping the health of companion animals and possible species-specific differences.

Re-evaluating the placebo response in recent canine dietary epilepsy trials

The placebo response is a common phenomenon. Limited evidence is available about its magnitude in canine epilepsy trials, even though it can significantly influence the efficacy evaluation of new treatments. It was hypothesised that the placebo response is diminished when epilepsy trials are conducted in a prospective crossover design. Seizure data spanning six months from three previous multicenter epilepsy studies were analysed. The monthly seizure frequency of 60 dogs diagnosed with idiopathic epilepsy was calculated, comparing baseline data with placebo treatment. Furthermore, differentiation was made between dogs randomised to the placebo group early (Phase 1: first 3 months) or later during the study (Phase 2: second 3 months).The analysis did not reveal any placebo response in terms of monthly seizure frequency. Instead, an increase was noted during the placebo treatment period, with a mean of 2.95 seizures per month compared to 2.30 seizures per month before study entry (p = 0.0378). Additionally, a notable phase effect was observed. Dogs receiving the placebo in the second study phase exhibited a significant increase in monthly seizure frequency compared to baseline (p = 0.0036). Conversely, no significant difference from baseline was observed for dogs receiving the placebo in the first study phase. These findings underscore the considerable variability in placebo responses observed in trials for canine epilepsy, contrasting with previous limited data. The identified phase effect should be carefully considered in the design and evaluation of canine epilepsy trials to ensure a more accurate assessment of efficacy for new treatments.

Feasibility of in-home electroencephalographic and actigraphy recordings in dogs

Introduction

Idiopathic epilepsy is a prevalent neurological disease in dogs. Dogs with epilepsy often present with behavioral comorbidities such as aggression, anxiety, and fear. These behaviors are consistent with pre, post, or interictal behaviors, prodromal changes, seizure-precipitating factors, or absence and focal seizures. The overlap in behavior presentations and lack of objective research methods for quantifying and classifying canine behavior makes determining the cause difficult. Behavioral comorbidities in addition to the task of caring for an epileptic animal have a significant negative impact on dog and caregiver quality of life.

Methods

This pilot study aimed to assess the feasibility of a novel technology combination for behavior classification and epileptic seizure detection for a minimum 24-h recording in the dog's home environment. It was expected that combining electroencephalography (EEG), actigraphy, and questionnaires would be feasible in the majority of trials. A convenience sample of 10 community-owned dogs was instrumented with wireless video-EEG and actigraphy for up to 48 h of recording at their caregiver's home. Three questionnaires (maximum 137 questions) were completed over the recording period by caregivers to describe their dog's everyday behavior and habits.

Results

Six of the 10 included dogs had combined EEG and actigraphy recordings for a minimum of 24 h.

Discussion

This shows that in-home EEG and actigraphy recordings are possible in community-owned dogs and provides a basis for a prospective study examining the same technology combination in a larger sample size.

Remote Monitoring of Canine Patients Treated for Pruritus during the COVID-19 Pandemic in Florida Using a 3-D Accelerometer

The medical management of chronic canine pruritic dermatologic conditions is challenging and often frustrating. This is a report that shows one way of aiding the management of pruritic dogs using a remote monitoring device. It is often difficult for veterinarians to get dog owners to return to the clinic once a dog is treated. It is possible that a 3-D accelerometer device could provide information to the clinic staff on the success or failure of a pruritus treatment plan while the dog was cared for at home. Eighty-seven dogs and their owners came to a Florida dermatology specialty clinic or its general practice hospital to be evaluated and treated for pruritus. An ANIMO® 3-D accelerometer was placed on the collar of dogs diagnosed and treated for pruritus. Dogs that completed this study were monitored for 120 days (4 months). The ANIMO smart phone application monitored a dog's daily scratching, shaking, sleeping, activity, and resting and summarized this information in a daily report visible on the pet owner's smart phone. An additional variable (grooming minutes per day) could be seen by the study team that was not yet available in the app. The use of a 3-D accelerometer enabled veterinarians to continuously monitor dogs at home when they were being treated for itching. Clinic staff kept in touch with the owners by phone and could change therapy or bring the dog back for a recheck if problems were seen. Daily reports were combined into line charts that showed plots of scratching, shaking, grooming, and sleeping over four months. Veterinarians were able to remotely monitor dogs that had been treated for pruritus for up to four months through use of a collar-borne monitoring device. Dog owners and clinic staff used the daily summaries accessible through a smart phone application. Dogs seemed to tolerate the device well because of its small size, light weight, long battery life, and unobtrusive nature.

Tools and techniques for classifying behaviours in canine epilepsy

Idiopathic epilepsy is the most common neurological disease in dogs. Similar to humans, dogs with epilepsy often experience behavioural comorbidities such as increased fear, anxiety, and aggression, as reported by their caregivers. Investigations of behaviour in canine epilepsy have yet to untangle interictal and pre and postictal behaviours, prodromal changes, and seizure-precipitating factors. Under-recognition of absence and focal seizures further complicates these assessments. These complex behavioural presentations in combination with caring for an epileptic animal have a significant negative impact on the dog's and caregiver's quality of life. Despite the growing recognition of behavioural comorbidities and their impact on quality of life in dogs with epilepsy, few objective research methods for classifying and quantifying canine behaviour exist. This narrative review examines the strengths, limitations, and granularity of three tools used in the investigation of canine behaviour and epilepsy; questionnaires, electroencephalography, and actigraphy. It suggests that a prospective combination of these three tools has the potential to offer improvements to the objective classification and quantification of canine behaviour in epilepsy.

Machine learning based canine posture estimation using inertial data

The aim of this study was to design a new canine posture estimation system specifically for working dogs. The system was composed of Inertial Measurement Units (IMUs) that are commercially available, and a supervised learning algorithm which was developed for different behaviours. Three IMUs, each containing a 3-axis accelerometer, gyroscope, and magnetometer, were attached to the dogs' chest, back, and neck. To build and test the model, data were collected during a video-recorded behaviour test where the trainee assistance dogs performed static postures (standing, sitting, lying down) and dynamic activities (walking, body shake). Advanced feature extraction techniques were employed for the first time in this field, including statistical, temporal, and spectral methods. The most important features for posture prediction were chosen using Select K Best with ANOVA F-value. The individual contributions of each IMU, sensor, and feature type were analysed using Select K Best scores and Random Forest feature importance. Results showed that the back and chest IMUs were more important than the neck IMU, and the accelerometers were more important than the gyroscopes. The addition of IMUs to the chest and back of dog harnesses is recommended to improve performance. Additionally, statistical and temporal feature domains were more important than spectral feature domains. Three novel cascade arrangements of Random Forest and Isolation Forest were fitted to the dataset. The best classifier achieved an f1-macro of 0.83 and an f1-weighted of 0.90 for the prediction of the five postures, demonstrating a better performance than previous studies. These results were attributed to the data collection methodology (number of subjects and observations, multiple IMUs, use of common working dog breeds) and novel machine learning techniques (advanced feature extraction, feature selection and modelling arrangements) employed. The dataset and code used are publicly available on Mendeley Data and GitHub, respectively.

A six-month prospective, randomised, double-blinded, placebo-controlled, crossover, dietary trial design to investigate the potential of psychobiotics on seizure semiology and comorbidities in canine epilepsy: study protocol

BACKGROUND

Epilepsy is the most common chronic neurological disease in dogs. More than two-thirds of these patients suffer from associated behavioural comorbidities. The latter could have their origin in partially overlapping pathomechanisms, with the intestinal microbiome as a potential key link between them. The current arsenal of drugs for epilepsy management remains limited. Most canine patients continue to have seizures despite treatment and the occurrence of comorbidities is not sufficiently addressed, limiting quality of life of affected dogs and owners. Therefore, novel additional epilepsy management options are urgently needed. The microbiome-gut-brain axis may serve as a new target for the development of innovative multimodal therapeutic approaches to overcome current shortcomings in epilepsy management.

METHODS

A six-month prospective, randomised, double-blinded, placebo-controlled, crossover, dietary trial was designed to investigate the potential of the psychobiotic Bifidobacterium longum on behavioural comorbidities in canine epilepsy. Seizure semiology will be evaluated as a secondary outcome measure. Thirty-four privately owned dogs are planned to be included in the ongoing study meeting the following inclusion criteria: Dogs displaying increased anxiety/fear behaviour since the start of the idiopathic epilepsy. Tier II confidence level of the International Veterinary Epilepsy Task Force for the diagnosis of idiopathic epilepsy, with a maximum seizure interval of 3 month and a minimum of three generalised seizures within that period and chronically treated with at least one antiseizure drug without improvement in seizure frequency Each dog will receive the allocated supplement (probiotic vs. placebo) alongside its normal diet for a 3-month period. After a three-week wash out period, the second phase starts by administering the respective other supplement for another 3 months.

DISCUSSION

The current study considers modern high-quality standards for epilepsy medication trials. Common biasing effects should be limited to a possible minimum (regression-to-the mean effect, placebo effect, observer effect), ensuring a high validity and accuracy of the acquired results, thus enabling a representative nature of the efficacy of Bifidobacterium longum as add-on supplement for dogs suffering from epilepsy and its comorbidities. This publication should provide a description of the study procedure and data acquisition methods, including prognosed statistical analysis.

Effect of an intervention of exercise on sleep and seizure frequency in idiopathic epileptic dogs

OBJECTIVE

The goal of this study was to compare sleep and seizure frequency between epileptic dogs prescribed a 20% activity increase and epileptic dogs not prescribed an activity increase.

METHODS

Sixty-nine dogs receiving anti-epileptic drug therapy were enrolled in a 6-month prospective, randomised, placebo-controlled clinical trial with an intention-to-treat analysis. A canine activity monitoring device was used to measure activity levels and sleep scores.

RESULTS

Using an intention-to-treat analysis, the treatment group had an average of 0.381 more seizures per month (95% CI: 0.09 to 0.68) compared with the control group, although the difference in seizure days per month was not statistically significant. In a subgroup analysis of dogs whose activity increased by at least 10%, partial compliers had 0.719 more seizures per month (95% CI: 0.22 to 1.22) and 0.581 seizure days per month (95% CI: 0.001 to 1.16) compared with the control group. Sleep scores increased by 1.2% in the treatment compared with the control group (95% CI: 0.2 to 2.3%).

CONCLUSIONS

Seizure frequency and sleep score increased slightly, but significantly, in dogs with idiopathic epilepsy prescribed an increase in activity, compared with a control group.

Novel subscalp and intracranial devices to wirelessly record and analyze continuous EEG in unsedated, behaving dogs in their natural environments: A new paradigm in canine epilepsy research

Epilepsy is characterized by unprovoked, recurrent seizures and is a common neurologic disorder in dogs and humans. Roughly 1/3 of canines and humans with epilepsy prove to be drug-resistant and continue to have sporadic seizures despite taking daily anti-seizure medications. The optimization of pharmacologic therapy is often limited by inaccurate seizure diaries and medication side effects. Electroencephalography (EEG) has long been a cornerstone of diagnosis and classification in human epilepsy, but because of several technical challenges has played a smaller clinical role in canine epilepsy. The interictal (between seizures) and ictal (seizure) EEG recorded from the epileptic mammalian brain shows characteristic electrophysiologic biomarkers that are very useful for clinical management. A fundamental engineering gap for both humans and canines with epilepsy has been the challenge of obtaining continuous long-term EEG in the patients' natural environment. We are now on the cusp of a revolution where continuous long-term EEG from behaving canines and humans will be available to guide clinicians in the diagnosis and optimal treatment of their patients. Here we review some of the devices that have recently emerged for obtaining long-term EEG in ambulatory subjects living in their natural environments.

Neurosurgery in canine epilepsy

Epilepsy surgery is functional neurosurgery applied to drug-resistant epilepsy. Although epilepsy surgery has been established and achieves fair to good outcomes in human medicine, it is still an underdeveloped area in veterinary medicine. With the spread of advanced imaging and neurosurgical modalities, intracranial surgery has become commonplace in the veterinary field, and, therefore, it is natural that expectations for epilepsy surgery increase. This review summarizes current standards of intracranial epilepsy surgery in human medicine and describes its current status and expectation in veterinary medicine. Intracranial epilepsy surgery is classified generally into resection surgery, represented by cortical resection, lobectomy, and lesionectomy, and disconnection surgery, such as corpus callosotomy and multiple subpial transection. In dogs with drug-resistant epilepsy, corpus callosotomy is available as a disconnection surgery for generalized epilepsy. However, other types of disconnection and resection surgeries for focal epilepsy are limited to experimental studies in laboratory dogs and/or anecdotal case reports of lesionectomy, such as tumor or encephalocele removal, without epileptogenic evidence. Veterinary epilepsy surgery is a new and challenging neurosurgery field; with the development of presurgical evaluations such as advanced electroencephalography and neuroimaging, it may become more readily practiced.

The Prospects of Non-EEG Seizure Detection Devices in Dogs

The unpredictable nature of seizures is challenging for caregivers of epileptic dogs, which calls the need for other management strategies such as seizure detection devices. Seizure detection devices are systems that rely on non-electroencephalographic (non-EEG) ictal changes, designed to detect seizures. The aim for its use in dogs would be to provide owners with a more complete history of their dog's seizures and to help install prompt (and potentially life-saving) intervention. Although seizure detection via wearable intracranial EEG recordings is associated with a higher sensitivity in humans, there is robust evidence for reliable detection of generalized tonic-clonic seizures (GTCS) using non-EEG devices. Promising non-EEG changes described in epileptic humans, include heart rate variability (HRV), accelerometry (ACM), electrodermal activity (EDA), and electromyography (EMG). Their sensitivity and false detection rate to detect seizures vary, however direct comparison of studies is nearly impossible, as there are many differences in study design and standards for testing. A way to improve sensitivity and decrease false-positive alarms is to combine the different parameters thereby profiting from the strengths of each one. Given the challenges of using EEG in veterinary clinical practice, non-EEG ictal changes could be a promising alternative to monitor seizures more objectively. This review summarizes various seizure detection devices described in the human literature, discusses their potential use and limitations in veterinary medicine and describes what is currently known in the veterinary literature.

Detection of Generalized Tonic–Clonic Seizures in Dogs With a Seizure Detection System Established Using Acceleration Data and the Mahalanobis Distance: A Preliminary Study

Caregivers of dogs with epilepsy experience severe stress due to unpredictable seizures. Hence, they feel the need for a better management strategy. A seizure detection system (SDS), which can identify seizures and provide notifications to caregivers immediately, is required to address this issue. The current study aimed to establish a wearable automatic SDS using acceleration data and the Mahalanobis distance and to preliminarily investigate its feasibility among dogs. A generalized tonic–clonic seizure (GTCS) was targeted because it is the most common type of seizure and can have serious consequences (i.e., status epilepticus). This study comprised three phases. First, the reference datasets of epileptic and non-epileptic activities were established using acceleration data of GTCSs in 3 dogs and daily activities in 27 dogs. Second, the GTCS-detecting algorithm was created using the reference datasets and was validated using other acceleration data of GTCSs in 4 epileptic dogs and daily activities in 27 dogs. Third, a feasibility test of the SDS prototype was performed in three dogs with epilepsy. The algorithm was effective in identifying all acceleration data of GTCSs as seizures and all acceleration data of daily activities as non-seizure activities. Dogs with epilepsy were monitored with the prototype for 48–72 h, and three GTCSs were identified. The prototype detected all GTCSs accurately. A false positive finding was not obtained unless the accelerometer was displaced. Hence, a method that can detect epileptic seizures, particularly GTCSs, was established. Nevertheless, further large-scale studies must be conducted before the method can be commercialized.

Owner's Perception of Seizure Detection Devices in Idiopathic Epileptic Dogs

Accurate knowledge of seizure frequency is key to optimising treatment. New methods for detecting epileptic seizures are currently investigated in humans, which rely on changes in biomarkers, also called seizure detection devices. Critical to device development, is understanding user needs and requirements. No information on this subject has been published in veterinary medicine. Many dog health collars are currently on the market, but none has proved to be a promising seizure detector. An online survey was created and consisted of 27 open, closed, and scaled questions divided over two parts: part one focused on general questions related to signalment and seizure semiology, the second part focused specifically on the use of seizure detection devices. Two hundred and thirty-one participants caring for a dog with idiopathic epilepsy, were included in the study. Open questions were coded using descriptive coding by two of the authors independently. Data was analysed using descriptive statistics and binary logistic regression. Our results showed that the unpredictability of seizures plays a major part in the management of canine epilepsy and dog owners have a strong desire to know when a seizure occurs. Nearly all dog owners made changes in their daily life, mainly focusing on intensifying supervision. Owners believed seizure detection devices would improve their dog's seizure management, including a better accuracy of seizure frequency and the ability to administer emergency drugs more readily. Owners that were already keeping track of their dog's seizures were 4.2 times more likely to show confidence in using seizure detection devices to manage their pet's seizures, highlighting the need for better monitoring systems. Our results show that there is a receptive market for wearable technology as a new management strategy in canine epilepsy and this topic should be further explored.

Evaluation of a collar-mounted accelerometer for detecting seizure activity in dogs

Background

The majority of dogs with idiopathic epilepsy continue to have seizures despite appropriate treatment.

Objectives

To assess the use of a commercially available, collar‐mounted accelerometer to detect generalized seizures in dogs.

Animals

Twenty two client‐owned dogs with idiopathic epilepsy.

Methods

Six‐month prospective clinical study during which dogs wore a collar‐mounted accelerometer. Seizure documentation was based on owner observations and video recordings. The accelerometer used a predefined algorithm to detect seizures in the first study phase, and an individualized algorithm in the second study phase. Caregivers completed a quality of life (QoL) questionnaire at the initial and final study visit.

Results

Using the predefined algorithm, the accelerometer detected seizures with a sensitivity of 18.6% (95% CI [13.4%, 23.8%]) and mean false detection rate of 0.096/day. Values did not change significantly with use of an individualized algorithm (sensitivity 22.1%, 95% CI [15.1%, 29.0%]; false detection rate 0.054/day). Mean composite QoL score was significantly improved at study completion (50.42) compared to study initiation (39.53; P = .005), and this change was moderately correlated with a change in weekly exercise (r = 0.46, P = .05).

Conclusions and Clinical Importance

Generalized seizures in dogs can be detected with a collar‐mounted accelerometer, but the overall sensitivity is low.