Elif Surer

Exergaming and rehabilitation: A methodology for the design of effective and safe therapeutic exergames☆

Abstract We present here a comprehensive definition of therapeutic exergames from which a methodology to create safe exergames for real therapy pathways is derived. Three main steps are identified. (I) A clear identification of all the exercise requirements, not only in terms of goals of the therapy, but also in terms of additional constraints. Characteristic parameters for determining the challenge level and to assess progression are also defined in this phase. (II) The exercise is transformed into a Virtual Exercise, in which all the exercise elements are implemented inside a simple virtual environment. In this step the discussion between clinical and ICT teams allows maximizing the effectiveness of exergames implementation. (III) The final exergame is realized by introducing on top of the exercise all the game elements suggested by good game design to maximize entertainment. A clear line between exercises and games is drawn here. We illustrate the methodology with exergames designed for (1) balance and posture and (2) neglect rehabilitation, implemented and tested with post-stroke patients training autonomously at home. The methodology can have a broader impact as it can be applied also in other gaming fields in which the requirements go beyond entertainment.

A markerless estimation of the ankle–foot complex 2D kinematics during stance

A markerless technique is proposed and applied to estimate the two-dimensional joint kinematics of the shank and foot complex during the stance phase. Image sequences were acquired with a single camera from three healthy subjects while walking barefoot and with socks. Automatic segmentation of the shank and foot was performed to isolate the moving body from the background. A multi-rigid body model for the shank and foot complex, with the relevant segment anatomical axes, was defined and an image cross-correlation technique was applied to detect the anatomical axes locations throughout the movement. The proposed markerless technique was validated by acquiring the same trials also with a stereophotogrammetric marker-based system and a simple marker set. Differences in the joint kinematics estimates obtained with the two techniques fall in most cases within the intra-subject variability showing that, in selected applications, the markerless technique may replace more expensive and more experimental time demanding marker-based techniques.

User Perspectives on Exergames Designed to Explore the Hemineglected Space for Stroke Patients With Visuospatial Neglect: Usability Study

Background Visuospatial neglect due to stroke is characterized by the inability to perceive stimuli emerging in the area opposite to the side of brain damage. Besides adopting conventional rehabilitation methods to treat neglect symptoms, the use of virtual reality (VR) is becoming increasingly popular. We designed a series of 9 exergames aimed to improve exploration of the neglected side of space. When new VR interventions are designed, it is important to assess the usability aspects of such management strategies within the target population. To date, most studies used questionnaires to assess user satisfaction with the intervention or product being tested. However, only a combination of both quantitative and qualitative data allows a full picture of user perspective. Objective The purpose of this study was to quantitatively and qualitatively assess patient and therapist perspectives of a VR intervention based on the series of 9 exergames designed to explore hemineglected space. Specifically, we wanted to evaluate (1) perceived-user friendliness of the exergames, (2) attitude towards using the exergames, and (3) intention to use the exergames in the future. Methods A total of 19 participants (7 patients, 12 therapists) evaluated the exergames they had used 5 times a week during 3 weeks. The Technology Acceptance Model (TAM) questionnaire was filled out after the intervention. Based on those responses, we conducted focus group interviews (with therapists) and individual interviews (with patients). To analyze the TAM questionnaires, we used descriptive statistics. We adopted content and comparative analysis to analyze the interviews and drew illustration maps to analyze the focus group interviews. Results The therapists took a more critical stance with a mean TAM questionnaire total score of 48.6 (SD 4.5) compared to the patients who had a mean total score of 56.1 (SD 12.3). The perceived user-friendliness score was 5.6 (SD 1.4) for patients and 4.9 (SD 1.4) for therapists. The attitude towards using the exergames was rated 4.8 (SD 1.9) by patients and 3.6 (SD 1.4) by therapists, respectively. The intention to use the exergames in the future was rated 3.9 (SD 2.1) by patients and 3.7 (SD 1.8) by therapists. We gained information on how to improve the exergames in the interviews. Conclusions Patients and therapists perceived the exergames as user-friendly; however, using the games further with the actual test version was not perceived as conceivable. The therapists were generally more critical towards future use than the patients. Therefore, involving both users to achieve acceptable and user-friendly versions of game-based rehabilitation for the future is deemed crucial and warranted. Trial Registration Clinicaltrials.gov NCT02353962; https://clinicaltrials.gov/ct2/show/NCT02353962 (Archived by WebCite at http://www.webcitation.org/6soxIJlAZ)

Algorithms Based on Computational Intelligence for Autonomous Physical Rehabilitation at Home

Exergames provide efficient and motivating training mechanisms to support physical rehabilitation at home. Nonetheless, current exergame examples lack some important aspects which cannot be disregarded in rehabilitation. Exergames should: (i) modify the game difficulty adapting to patient’s gameplay performance, (ii) monitor if the exercise is correctly executed, and (iii) provide continuous motivation. In this study, we present a game engine which implements computer intelligence-based solutions to provide real-time adaptation, on-line monitoring and an engaging gameplay experience. The game engine applies real-time adaptation using the Quest Bayesian approach to modify the game difficulty according to the patient’s performance. Besides, it employs a fuzzy system to monitor the execution of the exercises according to the parameters set by the therapists and provides on-line feedback to guide the patient during the execution of the exercise. Finally, a motivating game experience is provided using rewards and adding random enrichments during the game.

Methods and Technologies for Gait Analysis

Gait analysis is a highly active research area with a wide range of applications in clinical settings, surveillance and human-computer interaction. The focus of this chapter is the clinical aspect of gait analysis, in which accuracy and precision are essential. Subsequently, the chapter focuses on various techniques of measuring gait and introduces taxonomy for their analysis. From this perspective, motion measurements using motion capture and inertial sensors are presented. Motion capture techniques are analyzed under sections of marker-based and markerless techniques and their common applications are exemplified. Additionally, accelerometers, gyroscopes, magnetometers and their applications are presented in the inertial measurements section. Finally, force measurements and measurement of electrical activity of muscles are explained briefly.

Boosted multiple kernel learning for first-person activity recognition

Activity recognition from first-person (ego-centric) videos has recently gained attention due to the increasing ubiquity of the wearable cameras. There has been a surge of efforts adapting existing feature descriptors and designing new descriptors for the first-person videos. An effective activity recognition system requires selection and use of complementary features and appropriate kernels for each feature. In this study, we propose a data-driven framework for first-person activity recognition which effectively selects and combines features and their respective kernels during the training. Our experimental results show that use of Multiple Kernel Learning (MKL) and Boosted MKL in first-person activity recognition problem exhibits improved results in comparison to the state-of-the-art. In addition, these techniques enable the expansion of the framework with new features in an efficient and convenient way.

Exergames Encouraging Exploration of Hemineglected Space in Stroke Patients With Visuospatial Neglect: A Feasibility Study

Background Use of exergames can complement conventional therapy and increase the amount and intensity of visuospatial neglect (VSN) training. A series of 9 exergames—games based on therapeutic principles—aimed at improving exploration of the neglected space for patients with VSN symptoms poststroke was developed and tested for its feasibility. Objectives The goal was to determine the feasibility of the exergames with minimal supervision in terms of (1) implementation of the intervention, including adherence, attrition and safety, and (2) limited efficacy testing, aiming to document possible effects on VSN symptoms in a case series of patients early poststroke. Methods A total of 7 patients attended the 3-week exergames training program on a daily basis. Adherence of the patients was documented in a training diary. For attrition, the number of participants lost during the intervention was registered. Any adverse events related to the exergames intervention were noted to document safety. Changes in cognitive and spatial exploration skills were measured with the Zürich Maxi Mental Status Inventory and the Neglect Test. Additionally, we developed an Eye Tracker Neglect Test (ETNT) using an infrared camera to detect and measure neglect symptoms pre- and postintervention. Results The median was 14 out of 15 (93%) attended sessions, indicating that the adherence to the exergames training sessions was high. There were no adverse events and no drop-outs during the exergame intervention. The individual cognitive and spatial exploration skills slightly improved postintervention (P=.06 to P=.98) and continued improving at follow-up (P=.04 to P=.92) in 5 out of 7 (71%) patients. Calibration of the ETNT was rather error prone. The ETNT showed a trend for a slight median group improvement from 15 to 16 total located targets (+6%). Conclusions The high adherence rate and absence of adverse events showed that these exergames were feasible and safe for the participants. The results of the amount of exergames use is promising for future applications and warrants further investigations—for example, in the home setting of patients to augment training frequency and intensity. The preliminary results indicate the potential of these exergames to cause improvements in cognitive and spatial exploration skills over the course of training for stroke patients with VSN symptoms. Thus, these exergames are proposed as a motivating training tool to complement usual care. The ETNT showed to be a promising assessment for quantifying spatial exploration skills. However, further adaptations are needed, especially regarding calibration issues, before its use can be justified in a larger study sample.

Video-games based framework designed for the cognitive rehabilitation of children with Down Syndrome

Children with Down Syndrome have difficulties in solving cognitive tasks and there are special learning and rehabilitation programs to help them overcome these difficulties. In this study, a video-games based rehabilitation platform is developed so that children with Down Syndrome can continue their rehabilitation autonomously at home. For this purpose, a set of video games which aim to address cognitive tasks such as attention, memory, abstraction and finding relations are being designed and implemented. During the gameplay, measurements regarding brain activity and electro dermal activity are done and afterwards, a user-specific rehabilitation program is designed.

Voluntary behavior on cortical learning algorithm based agents

Operating autonomous agents inside a 3D workspace is a challenging problem domain in real-time for dynamic environments since it involves online interaction with ever-changing decision constraints. This study proposes a neuroscience inspired architecture to simulate autonomous agents with interaction capabilities inside a 3D virtual world. The environment stimulates the operating agents based on their place and course of action. They are expected to form a life cycle composed of behavior chunks inside this environment and continuously optimize it around the stimulated reward. The architecture is composed of specialized units that run Cortical Learning Algorithm (CLA) which models functional properties of layers II and III as in six layer theory of neocortex. This work focuses on extending it with functional properties of layers IV, V and basal ganglia to obtain voluntary behavior that is suitable for an autonomous agent. Through experimental scenarios, the architecture is observed and evaluated in order to obtain an apparent learning process. The communication between layers and internal connectivity of embedded CLA units are able to capture sequential and causal relations from the environment and the first evaluation of the implementation has high potential for future directions.

Physical and Cognitive Training of Children with Down Syndrome Using Video Games

In this study, a video-games based training platform that aims to provide user-specific physical and cognitive tasks is developed so that children with Down Syndrome can continue their training autonomously at home. For this purpose, a set of video games which addresses physical activities (balance and feet coordination) and cognitive tasks (abstraction, memory and word-forming) are being designed and implemented. During the gameplays, center of pressure, brain activity and electrodermal activity measurements are done to identify the specific needs of the child and to tailor a training programme that addresses these difficulties.