Calibration-Free Mobile Eye-Tracking Using Corneal Imaging.

In this paper, we present and evaluate a calibration-free mobile eye-traking system. The system's mobile device consists of three cameras: an IR eye camera, an RGB eye camera, and a front-scene RGB camera. The three cameras build a reliable corneal imaging system that is used to estimate the user's point of gaze continuously and reliably. The system auto-calibrates the device unobtrusively. Since the user is not required to follow any special instructions to calibrate the system, they can simply put on the eye tracker and start moving around using it. Deep learning algorithms together with 3D geometric computations were used to auto-calibrate the system per user. Once the model is built, a point-to-point transformation from the eye camera to the front camera is computed automatically by matching corneal and scene images, which allows the gaze point in the scene image to be estimated. The system was evaluated by users in real-life scenarios, indoors and outdoors. The average gaze error was 1.6∘ indoors and 1.69∘ outdoors, which is considered very good compared to state-of-the-art approaches.

Characterizing Movement Patterns of Older Individuals with T2D in Free-Living Environments Using Wearable Accelerometers.

(1) Background: Type 2 Diabetes (T2D) is associated with reduced muscle mass, strength, and function, leading to frailty. This study aims to analyze the movement patterns (MPs) of older individuals with T2D across varying levels of physical capacity (PC). (2) Methods: A cross-sectional study was conducted among individuals aged 60 or older with T2D. Participants (n = 103) were equipped with a blinded continuous glucose monitoring (CGM) system and an activity monitoring device for one week. PC tests were performed at the beginning and end of the week, and participants were categorized into three groups: low PC (LPC), medium PC (MPC), and normal PC (NPC). Group differences in MPs and physical activity were analyzed using non-parametric Kruskal-Wallis tests for both categorical and continuous variables. Dunn post-hoc statistical tests were subsequently carried out for pairwise comparisons. For data analysis, we utilized pandas, a Python-based data analysis tool, and conducted the statistical analyses using the scipy.stats package in Python. The significance level was set at p < 0.05. (3) Results: Participants in the LPC group showed lower medio-lateral acceleration and higher vertical and antero-posterior acceleration compared to the NPC group. LPC participants also had higher root mean square values (1.017 m/s2). Moreover, the LPC group spent less time performing in moderate to vigorous physical activity (MVPA) and had fewer daily steps than the MPC and NPC groups. (4) Conclusions: The LPC group exhibited distinct movement patterns and lower activity levels compared to the NPC group. This study is the first to characterize the MPs of older individuals with T2D in their free-living environment. Several accelerometer-derived features were identified that could differentiate between PC groups. This novel approach offers a manpower-free alternative to identify physical deterioration and detect low PC in individuals with T2D based on real free-living physical behavior.

Personalisation of a virtual gaming system for children with motor impairments: performance and usability.

PURPOSE: To demonstrate the potential role of virtual game personalisation for use as a therapeutic modality to improve upper extremity function in children with cerebral palsy (CP). METHODS: The study tested a convenience sample of 60 typically developing children (TD) aged 6-10 years and 20 children with CP aged 7-11 years. Children participated in a single 30-min session when they played the game in accuracy mode (virtual targets are hit as they become progressively larger or smaller) or dwell mode (virtual targets are hit when the users remains on them for progressively shorter or longer durations). These two modes can be played in conventional (non-personalised), personalised and with and without arm weights conditions; weights were used for the TD group in order to ensure that game play would be sufficiently challenging as to require personalisation. We measured performance variables (frequency of changes in game level difficulty and accuracy as measured by percent success of hitting the virtual targets) in each condition and usability variables (self-reported perceived effort and enjoyment). RESULTS: Comparisons between the usability of the conventional and personalised conditions among typically developing children showed that although children self-reported significantly more effort while playing the personalised game, the level of enjoyment remained high (no significant differences between conventional and personalised game play conditions). In addition, comparisons between playing the personalised game with and without weights by typically developing children, indicated that percent success was significantly higher for the game played without weights, suggesting that the system is sensitive to dynamic changes in performance. Comparisons between the TD and CP groups showed that when the game was played in personalised dwell mode (hovering over the target for several seconds) children with CP progressed significantly less quickly through different difficulty levels compared to typically developing children. In contrast, no significant differences were found in accuracy mode (immediate response on target hit), between the TD and CP groups in any of the experimental conditions. DISCUSSION: The personalised game approach was shown to be enjoyable for both groups of users and able to change the level of difficulty in real time. The results suggest that this approach to gaming can provide motor challenges while preserving a high level of enjoyment. CONCLUSION: Personalised virtual therapy shows promise as a tool for upper extremity therapy for children with motor impairment.Implications for RehabiliationIn recent years, there has been an increase in the use of assistive technologies including virtual gaming in the general area of health care and clinical practice.Virtual gaming provides an interactive, real-time experiences that are flexible and ecologically valid ways to improve specific cognitive and motor abilities.Personalisation of virtual games entails dynamic adaptation of the parameters in real time according to the user's functional level).The results have demonstrated that personalised virtual gaming is enjoyable and feasible for typically developing children and children with cerebral palsy.The results suggest that this approach to gaming can provide motor challenges while preserving a high level of enjoyment.

3D Gaze Estimation Using RGB-IR Cameras.

In this paper, we present a framework for 3D gaze estimation intended to identify the user's focus of attention in a corneal imaging system. The framework uses a headset that consists of three cameras, a scene camera and two eye cameras: an IR camera and an RGB camera. The IR camera is used to continuously and reliably track the pupil and the RGB camera is used to acquire corneal images of the same eye. Deep learning algorithms are trained to detect the pupil in IR and RGB images and to compute a per user 3D model of the eye in real time. Once the 3D model is built, the 3D gaze direction is computed starting from the eyeball center and passing through the pupil center to the outside world. This model can also be used to transform the pupil position detected in the IR image into its corresponding position in the RGB image and to detect the gaze direction in the corneal image. This technique circumvents the problem of pupil detection in RGB images, which is especially difficult and unreliable when the scene is reflected in the corneal images. In our approach, the auto-calibration process is transparent and unobtrusive. Users do not have to be instructed to look at specific objects to calibrate the eye tracker. They need only to act and gaze normally. The framework was evaluated in a user study in realistic settings and the results are promising. It achieved a very low 3D gaze error (2.12°) and very high accuracy in acquiring corneal images (intersection over union-IoU = 0.71). The framework may be used in a variety of real-world mobile scenarios (indoors, indoors near windows and outdoors) with high accuracy.

Usability of clinical decision support system as a facilitator for learning the assistive technology adaptation process.

The aim of this study was to evaluate the usability of Ontology Supported Computerized Assistive Technology Recommender (OSCAR), a Clinical Decision Support System (CDSS) for the assistive technology adaptation process, its impact on learning the matching process, and to determine the relationship between its usability and learnability. Two groups of expert and novice clinicians (total, n = 26) took part in this study. Each group filled out system usability scale (SUS) to evaluate OSCAR's usability. The novice group completed a learning questionnaire to assess OSCAR's effect on their ability to learn the matching process. Both groups rated OSCAR's usability as "very good", (M [SUS] = 80.7, SD = 11.6, median = 83.7) by the novices, and (M [SUS] = 81.2, SD = 6.8, median = 81.2) by the experts. The Mann-Whitney results indicated that no significant differences were found between the expert and novice groups in terms of OSCAR's usability. A significant positive correlation existed between the usability of OSCAR and the ability to learn the adaptation process (rs = 0.46, p = 0.04). Usability is an important factor in the acceptance of a system. The successful application of user-centered design principles during the development of OSCAR may serve as a case study that models the significant elements to be considered, theoretically and practically in developing other systems. Implications for Rehabilitation Creating a CDSS with a focus on its usability is an important factor for its acceptance by its users. Successful usability outcomes can impact the learning process of the subject matter in general, and the AT prescription process in particular. The successful application of User-Centered Design principles during the development of OSCAR may serve as a case study that models the significant elements to be considered, theoretically and practically. The study emphasizes the importance of close collaboration between the developers and the end users.

Building an effective ontology for assistive technology.

The aim of this paper was to develop and validate an ontology for one class of assistive technology (AT), namely physically controllable pointing devices, using the Delphi method. Six occupational therapists with AT expertise identified important items and categories to the pointing device prescription through a three-round, structured process consisting of responses to a series of questionnaires. The intraclass correlation coefficient (ICC) was used to assess the interrater reliability of items included in categories related to the pointing devices and to the user profile. During the first round, the ICC ranged from 0.19 to 0.97; this improved to ICCs ranging from 0.72 to 1.0 during the second round. A full consensus was reached by the experts during the final round which included 218 items, divided into five categories, for the pointing device list, and 168 items, divided into six categories, for the user profile list. This ontology is expected to help achieve a more systematic regulation of the AT field, leading to greater standardization and increased knowledge sharing.

Effectiveness of a Clinical Decision Support System for Pointing Device Prescription.

OBJECTIVE: We present a novel, knowledge-driven approach to prescription of pointing devices that uses the Ontology-Supported Computerized Assistive Technology Recommender (OSCAR), a clinical decision support system (CDSS). METHOD: Fifty-five occupational therapists were divided into four groups: two assistive technology (AT) expert groups and two novice groups. Novice Group 1 used the OSCAR CDSS for the prescription process, and Novice Group 2 used the conventional method. OSCAR's effectiveness and its impact on users were evaluated. RESULTS: The ability of Novice Group 1 to make suitable pointing device prescriptions was similar to that of the two expert groups and was significantly better than that of Novice Group 2. The system positively affected Novice Group 1's learning of the prescription process. CONCLUSION: The structure and organized framework for clinical reasoning of the OSCAR CDSS appear to enable occupational therapy practitioners inexperienced in AT to achieve performance levels comparable to those of experts.

Building an ontology for assistive technology using the Delphi method.

The aim of this study was to use the Delphi method to develop and validate an ontology for one class of assistive technology (AT), namely physically controllable pointing devices. A three-round, structured process consisting of responses to a series of questionnaires was used to identify items and categories of importance to pointing device prescription from six occupational therapists with AT expertise. The intraclass correlation coefficient (ICC) was used to assess the interrater reliability of items included in categories related to the pointing devices and to the user profile by the six experts during each Delphi round. During the first round, the ICC ranged from 0.19 to 0.97; this improved to ICCs ranging from 0.72 to 1.0 during the second round. A full consensus was reached by the experts during the final round which included 218 items, divided into five categories, for the pointing device list, and 168 items, divided into six categories, for the user profile list. This ontology is expected to help achieve a more systematic regulation of the AT field, leading to greater standardization and increased knowledge sharing, which may help novice clinicians to achieve a level of competence in matching AT to a person with special needs.

Onto-clust--a methodology for combining clustering analysis and ontological methods for identifying groups of comorbidities for developmental disorders.

Children with developmental disorders usually exhibit multiple developmental problems (comorbidities). Hence, such diagnosis needs to revolve on developmental disorder groups. Our objective is to systematically identify developmental disorder groups and represent them in an ontology. We developed a methodology that combines two methods (1) a literature-based ontology that we created, which represents developmental disorders and potential developmental disorder groups, and (2) clustering for detecting comorbid developmental disorders in patient data. The ontology is used to interpret and improve clustering results and the clustering results are used to validate the ontology and suggest directions for its development. We evaluated our methodology by applying it to data of 1175 patients from a child development clinic. We demonstrated that the ontology improves clustering results, bringing them closer to an expert generated gold-standard. We have shown that our methodology successfully combines an ontology with a clustering method to support systematic identification and representation of developmental disorder groups.