Movement Quantification in Neurological Diseases: Methods and Applications

Deep learning approaches for seizure video analysis: A review

Seizure events can manifest as transient disruptions in the control of movements which may be organized in distinct behavioral sequences, accompanied or not by other observable features such as altered facial expressions. The analysis of these clinical signs, referred to as semiology, is subject to observer variations when specialists evaluate video-recorded events in the clinical setting. To enhance the accuracy and consistency of evaluations, computer-aided video analysis of seizures has emerged as a natural avenue. In the field of medical applications, deep learning and computer vision approaches have driven substantial advancements. Historically, these approaches have been used for disease detection, classification, and prediction using diagnostic data; however, there has been limited exploration of their application in evaluating video-based motion detection in the clinical epileptology setting. While vision-based technologies do not aim to replace clinical expertise, they can significantly contribute to medical decision-making and patient care by providing quantitative evidence and decision support. Behavior monitoring tools offer several advantages such as providing objective information, detecting challenging-to-observe events, reducing documentation efforts, and extending assessment capabilities to areas with limited expertise. The main applications of these could be (1) improved seizure detection methods; (2) refined semiology analysis for predicting seizure type and cerebral localization. In this paper, we detail the foundation technologies used in vision-based systems in the analysis of seizure videos, highlighting their success in semiology detection and analysis, focusing on work published in the last 7 years. We systematically present these methods and indicate how the adoption of deep learning for the analysis of video recordings of seizures could be approached. Additionally, we illustrate how existing technologies can be interconnected through an integrated system for video-based semiology analysis. Each module can be customized and improved by adapting more accurate and robust deep learning approaches as these evolve. Finally, we discuss challenges and research directions for future studies.

A Novel Physical Mobility Task to Assess Freezers in Parkinson's Disease

Freezing of gait (FOG), one of the most disabling features of Parkinson's disease (PD), is a brief episodic absence or marked reduction in stride progression despite the intention to walk. Progressively more people who experience FOG restrict their walking and reduce their level of physical activity. The purpose of this study is to develop and validate a physical mobility task that induces freezing of gait in a controlled environment, employing known triggers of FOG episodes according to the literature. To validate the physical mobility tasks, we recruited 10 volunteers that suffered PD-associated freezing (60.6 ± 7.29 years-old) with new FOG-Q ranging from 12 to 26. The validation of the proposed method was carried out using inertial sensors and video recordings. All subjects were assessed during the OFF and ON medication states. The total number of FOG occurrences during data collection was 144. The proposed tasks were able to trigger 120 FOG episodes, while the TUG test caused 24. The Inertial Measurement Unit (IMU) with accelerometer and gyroscope could not only detect FOG episodes but also allowed us to visualize the three types of FOG: akinesia, festination and trembling in place.

Kinematic Studies of the Go/No-Go Task as a Dynamic Sensorimotor Inhibition Task for Assessment of Motor and Executive Function in Stroke Patients: An Exploratory Study in a Neurotypical Sample

Inhibition of reaching and grasping actions as an element of cognitive control and executive function is a vital component of sensorimotor behaviour that is often impaired in patients who have lost sensorimotor function following a stroke. To date, there are few kinematic studies detailing the fine spatial and temporal upper limb movements associated with the millisecond temporal trajectory of correct and incorrect responses to visually driven Go/No-Go reaching and grasping tasks. Therefore, we aimed to refine the behavioural measurement of correct and incorrect inhibitory motor responses in a Go/No-Go task for future quantification and personalized rehabilitation in older populations and those with acquired motor disorders, such as stroke. An exploratory study mapping the kinematic profiles of hand movements in neurotypical participants utilizing such a task was conducted using high-speed biological motion capture cameras, revealing both within and between subject differences in a sample of healthy participants. These kinematic profiles and differences are discussed in the context of better assessment of sensorimotor function impairment in stroke survivors.

Novel 3D video action recognition deep learning approach for near real time epileptic seizure classification

Seizure semiology is a well-established method to classify epileptic seizure types, but requires a significant amount of resources as long-term Video-EEG monitoring needs to be visually analyzed. Therefore, computer vision based diagnosis support tools are a promising approach. In this article, we utilize infrared (IR) and depth (3D) videos to show the feasibility of a 24/7 novel object and action recognition based deep learning (DL) monitoring system to differentiate between epileptic seizures in frontal lobe epilepsy (FLE), temporal lobe epilepsy (TLE) and non-epileptic events. Based on the largest 3Dvideo-EEG database in the world (115 seizures/+680,000 video-frames/427GB), we achieved a promising cross-subject validation f1-score of 0.833±0.061 for the 2 class (FLE vs. TLE) and 0.763 ± 0.083 for the 3 class (FLE vs. TLE vs. non-epileptic) case, from 2 s samples, with an automated semi-specialized depth (Acc.95.65%) and Mask R-CNN (Acc.96.52%) based cropping pipeline to pre-process the videos, enabling a near-real-time seizure type detection and classification tool. Our results demonstrate the feasibility of our novel DL approach to support 24/7 epilepsy monitoring, outperforming all previously published methods.

Wrist Movement Variability Assessment in Individuals with Parkinson's Disease

(1) Background: Parkinson’s disease (PD) is a neurodegenerative disorder represented by the progressive loss of dopamine-producing neurons, it decreases the individual’s motor functions and affects the execution of movements. There is a real need to include quantitative techniques and reliable methods to assess the evolution of PD. (2) Methods: This cross-sectional study assessed the variability of wrist RUD (radial and ulnar deviation) and FE (flexion and extension) movements measured by two pairs of capacitive sensors (PS25454 EPIC). The hypothesis was that PD patients have less variability in wrist movement execution than healthy individuals. The data was collected from 29 participants (age: 62.13 ± 9.7) with PD and 29 healthy individuals (60.70 ± 8). Subjects performed the experimental tasks at normal and fast speeds. Six features that captured the amplitude of the hand movements around two axes were estimated from the collected signals. (3) Results: The movement variability was greater for healthy individuals than for PD patients (p < 0.05). (4) Conclusion: The low variability seen in the PD group may indicate they execute wrist RUD and FE in a more restricted way. The variability analysis proposed here could be used as an indicator of patient progress in therapeutic programs and required changes in medication dosage.

Enabling AI and Robotic Coaches for Physical Rehabilitation Therapy: Iterative Design and Evaluation with Therapists and Post-stroke Survivors

Artificial intelligence (AI) and robotic coaches promise the improved engagement of patients on rehabilitation exercises through social interaction. While previous work explored the potential of automatically monitoring exercises for AI and robotic coaches, the deployment of these systems remains a challenge. Previous work described the lack of involving stakeholders to design such functionalities as one of the major causes. In this paper, we present our efforts on eliciting the detailed design specifications on how AI and robotic coaches could interact with and guide patient’s exercises in an effective and acceptable way with four therapists and five post-stroke survivors. Through iterative questionnaires and interviews, we found that both post-stroke survivors and therapists appreciated the potential benefits of AI and robotic coaches to achieve more systematic management and improve their self-efficacy and motivation on rehabilitation therapy. In addition, our evaluation sheds light on several practical concerns (e.g. a possible difficulty with the interaction for people with cognitive impairment, system failures, etc.). We discuss the value of early involvement of stakeholders and interactive techniques that complement system failures, but also support a personalized therapy session for the better deployment of AI and robotic exercise coaches.

Gait Characterization and Analysis of Hereditary Amyloidosis Associated with Transthyretin Patients: A Case Series

Hereditary amyloidosis associated with transthyretin (ATTRv), is a rare autosomal dominant disease characterized by length-dependent symmetric polyneuropathy that has gait impairment as one of its consequences. The gait pattern of V30M ATTRv amyloidosis patients has been described as similar to that of diabetic neuropathy, associated with steppage, but has never been quantitatively characterized. In this study we aim to characterize the gait pattern of patients with V30M ATTRv amyloidosis, thus providing information for a better understanding and potential for supporting diagnosis and disease progression evaluation. We present a case series in which we conducted two gait analyses, 18 months apart, of five V30M ATTRv amyloidosis patients using a 12-camera, marker based, optical system as well as six force platforms. Linear kinematics, ground reaction forces, and angular kinematics results are analyzed for all patients. All patients, except one, showed a delayed toe-off in the second assessment, as well as excessive pelvic rotation, hip extension and external transverse rotation and knee flexion (in stance and swing phases), along with reduced vertical and mediolateral ground reaction forces. The described gait anomalies are not clinically quantified; thus, gait analysis may contribute to the assessment of possible disease progression along with the clinical evaluation.

Performance Index for in Home Assessment of Motion Abilities in Ataxia Telangiectasia: A Pilot Study

Background. It has been shown in the very recent literature that human walking generates rhythmic motor patterns with hidden time harmonic structures that are represented (at the subject’s comfortable speed) by the occurrence of the golden ratio as the the ratio of the durations of specific walking gait subphases. Such harmonic proportions may be affected—partially or even totally destroyed—by several neurological and/or systemic disorders, thus drastically reducing the smooth, graceful, and melodic flow of movements and altering gait self-similarities. Aim. In this paper we aim at, preliminarily, showing the reliability of a technologically assisted methodology—performed with an easy to use wearable motion capture system—for the evaluation of motion abilities in Ataxia-Telangiectasia (AT), a rare infantile onset neurodegenerative disorder, whose typical neurological manifestations include progressive gait unbalance and the disturbance of motor coordination. Methods. Such an experimental methodology relies, for the first time, on the most recent accurate and objective outcome measures of gait recursivity and harmonicity and symmetry and double support subphase consistency, applied to three AT patients with different ranges of AT severity. Results. The quantification of the level of the distortions of harmonic temporal proportions is shown to include the qualitative evaluations of the three AT patients provided by clinicians. Conclusions. Easy to use wearable motion capture systems might be used to evaluate AT motion abilities through recursivity and harmonicity and symmetry (quantitative) outcome measures.

Generalized Finite-Length Fibonacci Sequences in Healthy and Pathological Human Walking: Comprehensively Assessing Recursivity, Asymmetry, Consistency, Self-Similarity, and Variability of Gaits

Healthy and pathological human walking are here interpreted, from a temporal point of view, by means of dynamics-on-graph concepts and generalized finite-length Fibonacci sequences. Such sequences, in their most general definition, concern two sets of eight specific time intervals for the newly defined composite gait cycle, which involves two specific couples of overlapping (left and right) gait cycles. The role of the golden ratio, whose occurrence has been experimentally found in the recent literature, is accordingly characterized, without resorting to complex tools from linear algebra. Gait recursivity, self-similarity, and asymmetry (including double support sub-phase consistency) are comprehensively captured. A new gait index, named Φ-bonacci gait number, and a new related experimental conjecture—concerning the position of the foot relative to the tibia—are concurrently proposed. Experimental results on healthy or pathological gaits support the theoretical derivations.

Supporting the Assessment of Hereditary Transthyretin Amyloidosis Patients Based On 3-D Gait Analysis and Machine Learning

Hereditary Transthyretin Amyloidosis (vATTR-V30M) is a rare and highly incapacitating sensorimotor neuropathy caused by an inherited mutation (Val30Met), which typically affects gait, among other symptoms. In this context, we investigated the possibility of using machine learning (ML) techniques to build a model(s) that can be used to support the detection of the Val30Met mutation (possibility of developing the disease), as well as symptom onset detection for the disease, given the gait characteristics of a person. These characteristics correspond to 24 gait parameters computed from 3-D body data, provided by a Kinect v2 camera, acquired from a person while walking towards the camera. To build the model(s), different ML algorithms were explored: k-nearest neighbors, decision tree, random forest, support vector machines (SVM), and multilayer perceptron. For a dataset corresponding to 66 subjects (25 healthy controls, 14 asymptomatic mutation carriers, and 27 patients) and several gait cycles per subject, we were able to obtain a model that distinguishes between controls and vATTR-V30M mutation carriers (with or without symptoms) with a mean accuracy of 92% (SVM). We also obtained a model that distinguishes between asymptomatic and symptomatic carriers with a mean accuracy of 98% (SVM). These results are very relevant, since this is the first study that proposes a ML approach to support vATTR-V30M patient assessment based on gait, being a promising foundation for the development of a computer-aided diagnosis tool to help clinicians in the identification and follow-up of this disease. Furthermore, the proposed method may also be used for other neuropathies.

SnapKi-An Inertial Easy-to-Adapt Wearable Textile Device for Movement Quantification of Neurological Patients

The development of wearable health systems has been the focus of many researchers who aim to find solutions in healthcare. Additionally, the large potential of textiles to integrate electronics, together with the comfort and usability they provide, has contributed to the development of smart garments in this area. In the field of neurological disorders with motor impairment, clinicians look for wearable devices that may provide quantification of movement symptoms. Neurological disorders affect different motion abilities thus requiring different needs in movement quantification. With this background we designed and developed an inertial textile-embedded wearable device that is adaptable to different movement-disorders quantification requirements. This adaptative device is composed of a low-power 9-axis inertial unit, a customised textile band and a web and Android cross application used for data collection, debug and calibration. The textile band comprises a snap buttons system that allows the attachment of the inertial unit, as well as its connection with the analog sensors through conductive textile. The resulting system is easily adaptable for quantification of multiple motor symptoms in different parts of the body, such as rigidity, tremor and bradykinesia assessments, gait analysis, among others. In our project, the system was applied for a specific use-case of wrist rigidity quantification during Deep Brain Stimulation surgeries, showing its high versatility and receiving very positive feedback from patients and doctors.

TTR-FAP Progression Evaluation Based on Gait Analysis Using a Single RGB-D Camera

Transthyretin Familial Amyloid Polyneuropathy (TTR-FAP) is a rare and disabling neurological disorder caused by a mutation of the transthyretin gene. One of the disease's characteristics that mostly affects patients' quality of life is its influence on locomotion, with a variable evolution timing. Quantitative motion analysis is useful for assessing motor function, including gait, in diseases affecting movement. However, it is still an evolving field, especially in TTR-FAP, with only a few available studies. A single markerless RGB-D camera provides 3-D body joint data in a less expensive, more portable and less intrusive way than reference multi-camera marker-based systems for motion capture. In this contribution, we investigate if a gait analysis system based on a RGB-D camera can be used to detect gait changes over time for a given TTR-FAP patient. 3-D data provided by that system and a reference system were acquired from six TTR-FAP patients, while performing a simple gait task, once and then a year and a half later. For each gait cycle and system, several gait parameters were computed. For each patient, we investigated if the RBG-D camera system is able to detect the existence or not of statistically significant differences between the two different acquisitions (separated by 1.5 years of disease evolution), in a similar way to the reference system. The obtained results show the potential of using a single RGB-D camera to detect relevant changes in spatiotemporal gait parameters (e.g., stride duration and stride length), during TTR-FAP patient follow-up.

Validation of a Single RGB-D Camera for Gait Assessment of Polyneuropathy Patients

Motion analysis systems based on a single markerless RGB-D camera are more suitable for clinical practice than multi-camera marker-based reference systems. Nevertheless, the validity of RGB-D cameras for motor function assessment in some diseases affecting gait, such as Transthyretin Familial Amyloid Polyneuropathy (TTR-FAP), is yet to be investigated. In this study, the agreement between the Kinect v2 and a reference system for obtaining spatiotemporal and kinematic gait parameters was evaluated in the context of TTR-FAP. 3-D body joint data provided by both systems were acquired from ten TTR-FAP symptomatic patients, while performing ten gait trials. For each gait cycle, we computed several spatiotemporal and kinematic gait parameters. We then determined, for each parameter, the Bland Altman’s bias and 95% limits of agreement, as well as the Pearson’s and concordance correlation coefficients, between systems. The obtained results show that an affordable, portable and non-invasive system based on an RGB-D camera can accurately obtain most of the studied gait parameters (excellent or good agreement for eleven spatiotemporal and one kinematic). This system can bring more objectivity to motor function assessment of polyneuropathy patients, potentially contributing to an improvement of TTR-FAP treatment and understanding, with great benefits to the patients’ quality of life.

Methodological considerations for kinematic analysis of upper limbs in healthy and poststroke adults Part II: a systematic review of motion capture systems and kinematic metrics

Background and purpose: To review the methods used to analyze the kinematics of upper limbs (ULs) of healthy and poststroke adults, namely the motion capture systems and kinematic metrics. Summary of review: A database of articles published in the last decade was compiled using the following search terms combinations: ("upper extremity" OR "upper limb" OR arm) AND (kinematic OR motion OR movement) AND (analysis OR assessment OR measurement). The articles included in this review: (1) had the purpose to analyze objectively three-dimension kinematics of ULs, (2) studied functional movements or activities of daily living involving ULs, and (3) studied healthy and/or poststroke adults. Fourteen articles were included (four studied a healthy sample, three analyzed poststroke patients, and seven examined both poststroke and healthy participants). Conclusion: Most articles used optoelectronic systems with markers; however, the presentation of laboratory and task-specific errors is missing. Markerless systems, used in some studies, seem to be promising alternatives for implementation of kinematic analysis in hospitals and clinics, but the literature proving their validity is scarce. Most articles analyzed "joint kinematics" and "end-point kinematics," mainly related with reaching. The different stroke locations of the samples were not considered in their analysis and only three articles described their psychometric properties. Implication of key findings: Future research should validate portable motion capture systems, document their specific error at the acquisition place and for the studied task, include grasping and manipulation analysis, and describe psychometric properties.

Understanding Patients' Behavior: Vision-Based Analysis of Seizure Disorders

A substantial proportion of patients with functional neurological disorders (FND) are being incorrectly diagnosed with epilepsy because their semiology resembles that of epileptic seizures (ES). Misdiagnosis may lead to unnecessary treatment and its associated complications. Diagnostic errors often result from an overreliance on specific clinical features. Furthermore, the lack of electrophysiological changes in patients with FND can also be seen in some forms of epilepsy, making diagnosis extremely challenging. Therefore, understanding semiology is an essential step for differentiating between ES and FND. Existing sensor-based and marker-based systems require physical contact with the body and are vulnerable to clinical situations such as patient positions, illumination changes, and motion discontinuities. Computer vision and deep learning are advancing to overcome these limitations encountered in the assessment of diseases and patient monitoring; however, they have not been investigated for seizure disorder scenarios. Here, we propose and compare two marker-free deep learning models, a landmark-based and a region-based model, both of which are capable of distinguishing between seizures from video recordings. We quantify semiology by using either a fusion of reference points and flow fields, or through the complete analysis of the body. Average leave-one-subject-out cross-validation accuracies for the landmark-based and region-based approaches of 68.1% and 79.6% in our dataset collected from 35 patients, reveal the benefit of video analytics to support automated identification of semiology in the challenging conditions of a hospital setting.

On the Fly Reporting of Human Body Movement based on Kinect v2

Human gait analysis is of utmost importance in understanding several aspects of human movement. In clinical practice, characterizing movement in order to obtain accurate and reliable information is a major challenge, and physicians usually rely on direct observation in order to evaluate a patient's motor abilities. In this contribution, a system that can objectively analyze the patients gait and generate an on the fly, targeted and optimized gait analysis report is presented. It is an extension to an existing system that could be used without interfering with the healthcare environment, which did not provide any on the fly feedback to physicians. Patient data are acquired using Kinect v2, followed by data processing, gait specific feature extraction, ending with the generation of a quantitative on the fly report. To the best of our knowledge, the complete system fills the gap as a proper gait analysis system, i.e., a low-cost tool that can be applied without interfering with the healthcare environment, provide quantitative gait information and on the fly feedback to physicians through a motion quantification report that can be useful in multiple areas.

A hierarchical multimodal system for motion analysis in patients with epilepsy

During seizures, a myriad of clinical manifestations may occur. The analysis of these signs, known as seizure semiology, gives clues to the underlying cerebral networks involved. When patients with drug-resistant epilepsy are monitored to assess their suitability for epilepsy surgery, semiology is a vital component to the presurgical evaluation. Specific patterns of facial movements, head motions, limb posturing and articulations, and hand and finger automatisms may be useful in distinguishing between mesial temporal lobe epilepsy (MTLE) and extratemporal lobe epilepsy (ETLE). However, this analysis is time-consuming and dependent on clinical experience and training. Given this limitation, an automated analysis of semiological patterns, i.e., detection, quantification, and recognition of body movement patterns, has the potential to help increase the diagnostic precision of localization. While a few single modal quantitative approaches are available to assess seizure semiology, the automated quantification of patients' behavior across multiple modalities has seen limited advances in the literature. This is largely due to multiple complicated variables commonly encountered in the clinical setting, such as analyzing subtle physical movements when the patient is covered or room lighting is inadequate. Semiology encompasses the stepwise/temporal progression of signs that is reflective of the integration of connected neuronal networks. Thus, single signs in isolation are far less informative. Taking this into account, here, we describe a novel modular, hierarchical, multimodal system that aims to detect and quantify semiologic signs recorded in 2D monitoring videos. Our approach can jointly learn semiologic features from facial, body, and hand motions based on computer vision and deep learning architectures. A dataset collected from an Australian quaternary referral epilepsy unit analyzing 161 seizures arising from the temporal (n = 90) and extratemporal (n = 71) brain regions has been used in our system to quantitatively classify these types of epilepsy according to the semiology detected. A leave-one-subject-out (LOSO) cross-validation of semiological patterns from the face, body, and hands reached classification accuracies ranging between 12% and 83.4%, 41.2% and 80.1%, and 32.8% and 69.3%, respectively. The proposed hierarchical multimodal system is a potential stepping-stone towards developing a fully automated semiology analysis system to support the assessment of epilepsy.

Deep facial analysis: A new phase I epilepsy evaluation using computer vision

Semiology observation and characterization play a major role in the presurgical evaluation of epilepsy. However, the interpretation of patient movements has subjective and intrinsic challenges. In this paper, we develop approaches to attempt to automatically extract and classify semiological patterns from facial expressions. We address limitations of existing computer-based analytical approaches of epilepsy monitoring, where facial movements have largely been ignored. This is an area that has seen limited advances in the literature. Inspired by recent advances in deep learning, we propose two deep learning models, landmark-based and region-based, to quantitatively identify changes in facial semiology in patients with mesial temporal lobe epilepsy (MTLE) from spontaneous expressions during phase I monitoring. A dataset has been collected from the Mater Advanced Epilepsy Unit (Brisbane, Australia) and is used to evaluate our proposed approach. Our experiments show that a landmark-based approach achieves promising results in analyzing facial semiology, where movements can be effectively marked and tracked when there is a frontal face on visualization. However, the region-based counterpart with spatiotemporal features achieves more accurate results when confronted with extreme head positions. A multifold cross-validation of the region-based approach exhibited an average test accuracy of 95.19% and an average AUC of 0.98 of the ROC curve. Conversely, a leave-one-subject-out cross-validation scheme for the same approach reveals a reduction in accuracy for the model as it is affected by data limitations and achieves an average test accuracy of 50.85%. Overall, the proposed deep learning models have shown promise in quantifying ictal facial movements in patients with MTLE. In turn, this may serve to enhance the automated presurgical epilepsy evaluation by allowing for standardization, mitigating bias, and assessing key features. The computer-aided diagnosis may help to support clinical decision-making and prevent erroneous localization and surgery.

The first Transthyretin Familial Amyloid Polyneuropathy gait quantification study - preliminary results

Transthyretin Familial Amyloid Polyneuropathy (TTR-FAP) is a rare neurological disease caused by a genetic mutation with a variable presentation and consequent challenging diagnosis, complex follow-up and treatment. At this moment, this condition has no cure and treatment options are under development. One of the disease's implications is a definite and progressive motor impairment that from the early stages compromises walking ability and daily life activities. The detection of this impairment is key for the disease onset diagnosis. With the goal of improving diagnosis of the symptoms and patients' quality of life, the authors have assessed the gait characteristics of subjects suffering from this condition. This contribution shows the results of a preliminary study, using a non-intrusive, markerless vision-based gait analysis tool. To the best of our knowledge, the reported results constitute the first gait analysis data of TTR-FAP mutation carriers.

A novel portable, low-cost kinect-based system for motion analysis in neurological diseases

Many neurological diseases, such as Parkinson's disease and epilepsy, can significantly impair the motor function of the patients, often leading to a dramatic loss of their quality of life. Human motion analysis is regarded as fundamental towards an early diagnosis and enhanced follow-up in this type of diseases. In this contribution, we present NeuroKinect, a novel system designed for motion analysis in neurological diseases. This system includes an RGB-D camera (Microsoft Kinect) and two integrated software applications, KiT (KinecTracker) and KiMA (Kinect Motion Analyzer). The applications enable the preview, acquisition, review and management of data provided by the sensor, which are then used for motion analysis of relevant events. NeuroKinect is a portable, low-cost and markerless solution that is suitable for use in the clinical environment. Furthermore, it is able to provide quantitative support to the clinical assessment of different neurological diseases with movement impairments, as demonstrated by its usage in two different clinical routine scenarios: gait analysis in Parkinson's disease and seizure semiology analysis in epilepsy.