Machine learning algorithms predict canine structural epilepsy with high accuracy.

Introduction: Clinical reasoning in veterinary medicine is often based on clinicians' personal experience in combination with information derived from publications describing cohorts of patients. Studies on the use of scientific methods for patient individual decision making are largely lacking. This applies to the prediction of the individual underlying pathology in seizuring dogs as well. The aim of this study was to apply machine learning to the prediction of the risk of structural epilepsy in dogs with seizures. Materials and methods: Dogs with a history of seizures were retrospectively as well as prospectively included. Data about clinical history, neurological examination, diagnostic tests performed as well as the final diagnosis were collected. For data analysis, the Bayesian Network and Random Forest algorithms were used. A total of 33 features for Random Forest and 17 for Bayesian Network were available for analysis. The following four feature selection methods were applied to select features for further analysis: Permutation Importance, Forward Selection, Random Selection and Expert Opinion. The two algorithms Bayesian Network and Random Forest were trained to predict structural epilepsy using the selected features. Results: A total of 328 dogs of 119 different breeds were identified retrospectively between January 2017 and June 2021, of which 33.2% were diagnosed with structural epilepsy. An overall of 89,848 models were trained. The Bayesian Network in combination with the Random feature selection performed best. It was able to predict structural epilepsy with an accuracy of 0.969 (sensitivity: 0.857, specificity: 1.000) among all dogs with seizures using the following features: age at first seizure, cluster seizures, seizure in last 24 h, seizure in last 6 month, and seizure in last year. Conclusion: Machine learning algorithms such as Bayesian Networks and Random Forests identify dogs with structural epilepsy with a high sensitivity and specificity. This information could provide some guidance to clinicians and pet owners in their clinical decision-making process.

Clinical findings and outcome in feline tetanus: a multicentric retrospective study of 27 cases and review of the literature.

Tetanus is a toxigenic illness caused by the action of Clostridium tetani neurotoxin (TeNT), which results in partial or generalized muscle stiffness in infected mammals and birds. The disease is rarely reported in cats due to their innate resistance to the toxin. This multicentric retrospective study aimed to describe a significant population of cats with a diagnosis of tetanus and report their signalment, clinical and neurological signs, diagnostic findings, treatment, and outcome. A retrospective search through medical records from 11 referral centers in Europe resulted in the identification of 27 cases of feline tetanus from July 2005 to April 2023. These cases were further compared with previously reported cases in the veterinary literature. Young cats were more commonly represented than older cats, with a median age of 4 years. Clinical signs were initially characterized by a lame and/or stiff limb, near the primary injury site, in 17/26 (65%) cats. Signs were focal or multifocal in 21/27 (78%) cats of this study and one typical sign was the inability to flex the most severely affected limbs. Electrodiagnostic studies revealed characteristic changes, such as continuous spontaneous motor unit discharges in both agonist and antagonist muscles. Such studies are particularly useful in focal and multifocal cases and should be performed to further support the diagnosis. The toxin was successfully identified in one case using the mouse bioassay. Treatment included antibiotherapy (metronidazole) in most cases, muscle relaxants, appropriate nursing cares and handling of potential complications. Overall, the outcome appeared to be positive, with only 1/27 (3.7%) cats being euthanized due to financial restrains. 23/25 (92%) cats returned to an independent ambulatory capacity on all limbs within a median delay of 25 days. Mild to moderate long-term sequelae were reported in eight (30%) cats. This multicentric study is the first to bring together such a large number of cats affected with tetanus. Presentation of the disease in cats differs from that observed in humans and dogs, with most cats being locally affected. Compared to previous reports of tetanus, this series of cats had a better outcome overall, especially for cats affected with generalized tetanus.

Client's understanding of instructions for small animals in a veterinary neurological referral center.

BACKGROUND: It is not known how much information clients retrieve from discharge instructions. OBJECTIVE: To investigate client's understanding of discharge instructions and influencing factors. ANIMALS: Dogs and cats being hospitalized for neurological diseases. METHODS: Clients were presented questionnaires regarding their pet's disease, diagnostics, treatments, prognosis and discharge instructions at time of discharge and 2 weeks later. The same questions were answered by discharging veterinarians at time of discharge. Clients answered additional questions regarding the subjective feelings during discharge conversation. Data collected included: data describing discharging veterinarian (age, gender, years of clinical experience, specialist status), data describing the client (age, gender, educational status). Raw percentage of agreement (RPA) between answers of clinicians and clients as well as factors potentially influencing the RPA were evaluated. RESULTS: Of 230 clients being approached 151 (65.7%) and 70 (30.4%) clients responded to the first and second questionnaire, respectively (130 dog and 30 cat owners). The general RPA between clinician's and client's responses over all questions together was 68.9% and 66.8% at the 2 time points. Questions regarding adverse effects of medication (29.0%), residual clinical signs (35.8%), and confinement instructions (36.8%) had the lowest RPAs at the first time point. The age of clients (P = .008) negatively influenced RPAs, with clients older than 50 years having lower RPA. CONCLUSIONS AND CLINICAL IMPORTANCE: Clients can only partially reproduce information provided at discharge. Only clients' increasing age influenced recall of information. Instructions deemed to be important should be specifically stressed during discharge.

Prevalence of nonconvulsive seizures and nonconvulsive status epilepticus in dogs and cats with a history of cluster seizures: A retrospective study.

BACKGROUND: Nonconvulsive seizures (NCS) and nonconvulsive status epilepticus (NCSE) are frequently observed in human patients. Diagnosis of NCS and NCSE only can be achieved by the use of electroencephalography (EEG). Electroencephalographic monitoring is rare in veterinary medicine and consequently there is limited data on frequency of NCS and NCSE. OBJECTIVES: Determine the prevalence of NCS and NCSE in dogs and cats with a history of cluster seizures. ANIMALS: Twenty-six dogs and 12 cats. METHODS: Retrospective study. Medical records of dogs and cats with cluster seizures were reviewed. Electroencephalography was performed in order to identify electrographic seizure activity after the apparent cessation of convulsive seizure activity. RESULTS: Nonconvulsive seizures were detected in 9 dogs and 2 cats out of the 38 patients (29%). Nonconvulsive status epilepticus was detected in 4 dogs and 2 cats (16%). Five patients had both NCS and NCSE. A decreased level of consciousness was evident in 6/11 patients with NCS, 3/6 also had NCSE. Mortality rate for patients with NCS (73%) and NCSE (67%) was much higher than that for patients with no seizure activity on EEG (27%). CONCLUSION AND CLINICAL IMPORTANCE: Prevalence of NCS and NCSE is high in dogs and cats with a history of cluster seizures. Nonconvulsive seizures and NCSE are difficult to detect clinically and are associated with higher in hospital mortality rates. Results indicate that prompt EEG monitoring should be performed in dogs and cats with cluster seizures.

Application accuracy of a frameless optical neuronavigation system as a guide for craniotomies in dogs.

BACKGROUND: Optical neuronavigation systems using infrared light to create a virtual reality image of the brain allow the surgeon to track instruments in real time. Due to the high vulnerability of the brain, neurosurgical interventions must be performed with a high precision. The aim of the experimental cadaveric study was to determine the application accuracy of a frameless optical neuronavigation system as guide for craniotomies by determining the target point deviation of predefined target points at the skull surface in the area of access to the cerebrum, cerebellum and the pituitary fossa. On each of the five canine cadaver heads ten target points were marked in a preoperative computed tomography (CT) scan. These target points were found on the cadaver skulls using the optical neuronavigation system. Then a small drill hole (1.5 mm) was drilled at these points. Subsequently, another CT scan was made. Both CT data sets were fused into the neuronavigation software, and the actual target point coordinates were identified. The target point deviation was determined as the difference between the planned and drilled target point coordinates. The calculated deviation was compared between two observers. RESULTS: The analysis of the target point accuracies of all dogs in both observers taken together showed a median target point deviation of 1.57 mm (range: 0.42 to 5.14 mm). No significant differences were found between the observers or the different areas of target regions. CONCLUSION: The application accuracy of the described system is similar to the accuracy of other optical neuronavigation systems previously described in veterinary medicine, in which mean values of 1.79 to 4.3 mm and median target point deviations of 0.79 to 3.53 mm were determined.

Case Report: Clinical Use of a Patient-Individual Magnetic Resonance Imaging-Based Stereotactic Navigation Device for Brain Biopsies in Three Dogs.

Three-dimensional (3D) printing techniques for patient-individual medicine has found its way into veterinary neurosurgery. Because of the high accuracy of 3D printed specific neurosurgical navigation devices, it seems to be a safe and reliable option to use patient-individual constructions for sampling brain tissue. Due to the complexity and vulnerability of the brain a particularly precise and safe procedure is required. In a recent cadaver study a better accuracy for the 3D printed MRI-based patient individual stereotactic brain biopsy device for dogs is determined compared to the accuracies of other biopsy systems which are currently used in veterinary medicine. This case report describes the clinical use of this 3D printed MRI-based patient individual brain biopsy device for brain sampling in three dogs. The system was characterized by a simple handling. Furthermore, it was an effective and reliable tool to gain diagnostic brain biopsy samples in dogs with no significant side effects.

Accuracy of a magnetic resonance imaging-based 3D printed stereotactic brain biopsy device in dogs.

BACKGROUND: Brain biopsy of intracranial lesions is often necessary to determine specific therapy. The cost of the currently used stereotactic rigid frame and optical tracking systems for brain biopsy in dogs is often prohibitive or accuracy is not sufficient for all types of lesion. OBJECTIVES: To evaluate the application accuracy of an inexpensive magnetic resonance imaging-based personalized, 3D printed brain biopsy device. ANIMALS: Twenty-two dog heads from cadavers were separated into 2 groups according to body weight (<15 kg, >20 kg). METHODS: Experimental study. Two target points in each cadaver head were used (target point 1: caudate nucleus, target point 2: piriform lobe). Comparison between groups was performed using the independent Student's t test or the nonparametric Mann-Whitney U Test. RESULTS: The total median target point deviation was 0.83 mm (range 0.09-2.76 mm). The separate median target point deviations for target points 1 and 2 in all dogs were 0.57 mm (range: 0.09-1.25 mm) and 0.85 mm (range: 0.14-2.76 mm), respectively. CONCLUSION AND CLINICAL IMPORTANCE: This magnetic resonance imaging-based 3D printed stereotactic brain biopsy device achieved an application accuracy that was better than the accuracy of most brain biopsy systems that are currently used in veterinary medicine. The device can be applied to every size and shape of skull and allows precise positioning of brain biopsy needles in dogs.

A concept for a 3D-printed patient-specific stereotaxy platform for brain biopsy -a canine cadaver study.

The treatment of intracranial lesions requires a precise diagnosis with subsequent identification of an adequate therapeutic approach. Stereotactic tumor biopsy may be considered the safest neurosurgical procedure in terms of anticipated results and potential surgical complications. The aim of the present paper was to demonstrate a new method of stereotactic biopsy, based on a patient-specific 3D printed platform in dogs. The system was tested on two canine cadavers, a small (Shih Tzu) and a large (Labrador) breed. Imaginary biopsy targets were defined in a superficial (caudate nucleus) and a deep (piriform lobe) position. Based on 3 Tesla MRI, individualized stereotactic platforms were designed using a semi-automatic approach, and manufactured additively using ABS M30. A pre- and intra-operative CT was performed to compare the planned vs. the realized needle position for precision analyses of the procedure. The target points varied with a precision between 0.09 mm and 0.48 mm. Manufacturing time required 480 to 700 min per platform. The presented patient-specific stereotactic system seems a suitable instrument for application in small animal neurosurgery. In particular, the implementation of relevant stereotactic data may help performing the procedure in rapid sequence and with higher precision than currently-used systems. Required adjustments and adaptions to the respective anatomical conditions are omitted and make the procedure reliable and safe.