A Care Knowledge Management System Based on an Ontological Model of Caring for People With Dementia: Knowledge Representation and Development Study.

BACKGROUND: Caregivers of people with dementia find it extremely difficult to choose the best care method because of complex environments and the variable symptoms of dementia. To alleviate this care burden, interventions have been proposed that use computer- or web-based applications. For example, an automatic diagnosis of the condition can improve the well-being of both the person with dementia and the caregiver. Other interventions support the individual with dementia in living independently. OBJECTIVE: The aim of this study was to develop an ontology-based care knowledge management system for people with dementia that will provide caregivers with a care guide suited to the environment and to the individual patient's symptoms. This should also enable knowledge sharing among caregivers. METHODS: To build the care knowledge model, we reviewed existing ontologies that contain concepts and knowledge descriptions relating to the care of those with dementia, and we considered dementia care manuals. The basic concepts of the care ontology were confirmed by experts in Korea. To infer the different care methods required for the individual dementia patient, the reasoning rules as defined in Semantic Web Rule Languages and Prolog were utilized. The accuracy of the care knowledge in the ontological model and the usability of the proposed system were evaluated by using the Pellet reasoner and OntOlogy Pitfall Scanner!, and a survey and interviews were conducted with caregivers working in care centers in Korea. RESULTS: The care knowledge model contains six top-level concepts: care knowledge, task, assessment, person, environment, and medical knowledge. Based on this ontological model of dementia care, caregivers at a dementia care facility in Korea were able to access the care knowledge easily through a graphical user interface. The evaluation by the care experts showed that the system contained accurate care knowledge and a level of assessment comparable to normal assessment tools. CONCLUSIONS: In this study, we developed a care knowledge system that can provide caregivers with care guides suited to individuals with dementia. We anticipate that the system could reduce the workload of caregivers.

Fall Detection for the Elderly Based on 3-Axis Accelerometer and Depth Sensor Fusion with Random Forest Classifier.

In this paper, we propose a new fall detection method that combines 3-axis accelerometer and depth sensors. By combining vision and acceleration-derived features we managed to minimize the false detection rate that is considerably higher when the decision is based on just one type of feature. Also, using machine learning has led to good generalization performance. In addition, we newly created fall database that are more realistic than previous ones. Experiment results show that the proposed method can efficiently detect falls.

Social skills training for children with autism spectrum disorder using a robotic behavioral intervention system.

We designed a robot system that assisted in behavioral intervention programs of children with autism spectrum disorder (ASD). The eight-session intervention program was based on the discrete trial teaching protocol and focused on two basic social skills: eye contact and facial emotion recognition. The robotic interactions occurred in four modules: training element query, recognition of human activity, coping-mode selection, and follow-up action. Children with ASD who were between 4 and 7 years old and who had verbal IQ ≥ 60 were recruited and randomly assigned to the treatment group (TG, n = 8, 5.75 ± 0.89 years) or control group (CG, n = 7; 6.32 ± 1.23 years). The therapeutic robot facilitated the treatment intervention in the TG, and the human assistant facilitated the treatment intervention in the CG. The intervention procedures were identical in both groups. The primary outcome measures included parent-completed questionnaires, the Autism Diagnostic Observation Schedule (ADOS), and frequency of eye contact, which was measured with the partial interval recording method. After completing treatment, the eye contact percentages were significantly increased in both groups. For facial emotion recognition, the percentages of correct answers were increased in similar patterns in both groups compared to baseline (P > 0.05), with no difference between the TG and CG (P > 0.05). The subjects' ability to play, general behavioral and emotional symptoms were significantly diminished after treatment (p < 0.05). These results showed that the robot-facilitated and human-facilitated behavioral interventions had similar positive effects on eye contact and facial emotion recognition, which suggested that robots are useful mediators of social skills training for children with ASD. Autism Res 2017,. © 2017 International Society for Autism Research, Wiley Periodicals, Inc. Autism Res 2017, 10: 1306-1323. © 2017 International Society for Autism Research, Wiley Periodicals, Inc.

Region-Based Object Recognition by Color Segmentation Using a Simplified PCNN.

In this paper, we propose a region-based object recognition (RBOR) method to identify objects from complex real-world scenes. First, the proposed method performs color image segmentation by a simplified pulse-coupled neural network (SPCNN) for the object model image and test image, and then conducts a region-based matching between them. Hence, we name it as RBOR with SPCNN (SPCNN-RBOR). Hereinto, the values of SPCNN parameters are automatically set by our previously proposed method in terms of each object model. In order to reduce various light intensity effects and take advantage of SPCNN high resolution on low intensities for achieving optimized color segmentation, a transformation integrating normalized Red Green Blue (RGB) with opponent color spaces is introduced. A novel image segmentation strategy is suggested to group the pixels firing synchronously throughout all the transformed channels of an image. Based on the segmentation results, a series of adaptive thresholds, which is adjustable according to the specific object model is employed to remove outlier region blobs, form potential clusters, and refine the clusters in test images. The proposed SPCNN-RBOR method overcomes the drawback of feature-based methods that inevitably includes background information into local invariant feature descriptors when keypoints locate near object boundaries. A large number of experiments have proved that the proposed SPCNN-RBOR method is robust for diverse complex variations, even under partial occlusion and highly cluttered environments. In addition, the SPCNN-RBOR method works well in not only identifying textured objects, but also in less-textured ones, which significantly outperforms the current feature-based methods.

Fast and scalable approximate spectral matching for higher order graph matching.

This paper presents a fast and efficient computational approach to higher order spectral graph matching. Exploiting the redundancy in a tensor representing the affinity between feature points, we approximate the affinity tensor with the linear combination of Kronecker products between bases and index tensors. The bases and index tensors are highly compressed representations of the approximated affinity tensor, requiring much smaller memory than in previous methods, which store the full affinity tensor. We compute the principal eigenvector of the approximated affinity tensor using the small bases and index tensors without explicitly storing the approximated tensor. To compensate for the loss of matching accuracy by the approximation, we also adopt and incorporate a marginalization scheme that maps a higher order tensor to matrix as well as a one-to-one mapping constraint into the eigenvector computation process. The experimental results show that the proposed method is faster and requires smaller memory than the existing methods with little or no loss of accuracy.

A new automatic parameter setting method of a simplified PCNN for image segmentation.

An automatic parameter setting method of a simplified pulse coupled neural network (SPCNN) is proposed here. Our method successfully determines all the adjustable parameters in SPCNN and does not need any training and trials as required by previous methods. In order to achieve this goal, we try to derive the general formulae of dynamic threshold and internal activity of the SPCNN according to the dynamic properties of neurons, and then deduce the sub-intensity range expression of each segment based on the general formulae. Besides, we extract information from an input image, such as the standard deviation and the optimal histogram threshold of the image, and attempt to build a direct relation between the dynamic properties of neurons and the static properties of each input image. Finally, the experimental segmentation results of the gray natural images from the Berkeley Segmentation Dataset, rather than synthetic images, prove the validity and efficiency of our proposed automatic parameter setting method of SPCNN.