A novel smart somatosensory wearable assistive device for older adults' home rehabilitation during the COVID-19 pandemic.

Background: Due to the Coronavirus disease 19 (COVID-19) related social distancing measures and health service suspension, physical activity has declined, leading to increased falling risk and disability, and consequently, compromising the older adult health. How to improve the quality of older adult life has become a crucial social issue. Objective: In traditional rehabilitation, manual and repetitive muscle training cannot identify the patient's rehabilitation effect, and increasing the willingness to use it is not easy. Therefore, based on the usability perspective, this study aims to develop a novel smart somatosensory wearable assistive device (called SSWAD) combined with wireless surface electromyography (sEMG) and exergame software and hardware technology. The older adult can do knee extension, ankle dorsiflexion, and ankle plantar flexion rehabilitation exercises at home. Meanwhile, sEMG values can be digitally recorded to assist physicians (or professionals) in judgment, treatment, or diagnosis. Methods: To explore whether the novel SSWAD could improve the older adult willingness to use and motivation for home rehabilitation, 25 frail older adult (12 males and 13 females with an average age of 69.3) perform the rehabilitation program with the SSWAD, followed by completing the system usability scale (SUS) questionnaire and the semi-structured interview for the quantitative and qualitative analyses. In addition, we further investigate whether the factor of gender or prior rehabilitation experience would affect the home rehabilitation willingness or not. Results: According to the overall SUS score, the novel SSWAD has good overall usability performance (77.70), meaning that the SSWAD makes older adult feel interested and improves their willingness for continuous rehabilitation at home. In addition, the individual item scores of SUS are shown that female older adult with prior rehabilitation experience perform better in "Learnability" (t = 2.35, p = 0.03) and "Confidence" (t = -3.24, p = 0.01). On the contrary, male older adult without rehabilitation experience are more willing to adopt new technologies (t = -2.73, p = 0.02), and perform better in "Learnability" (t = 2.18, p = 0.04) and "Confidence" (t = -3.75, p < 0.001) with the SSWAD. In addition, the result of the semi-structured interview shows that the operation of the SSWAD is highly flexible, thus reducing older adult burden during the rehabilitation exercise and using them long-term. Conclusion: This novel SSWAD receives consistently positive feedback regardless of the gender or prior rehabilitation experience of elders. The SSWAD could be used as a novel way of home rehabilitation for elders, especially during the COVID-19 pandemic. Older adult can do rehabilitation exercises at home, and physicians could make proper judgments or adjust suitable treatments online according to the sEMG data, which older adult can know their rehabilitation progress at the same time. Most importantly, older adult do not have to go to the hospital every time for rehabilitation, which significantly reduces time and the risk of infection.

Convolutional Neural Networks to Classify Alzheimer's Disease Severity Based on SPECT Images: A Comparative Study.

Image recognition and neuroimaging are increasingly being used to understand the progression of Alzheimer's disease (AD). However, image data from single-photon emission computed tomography (SPECT) are limited. Medical image analysis requires large, labeled training datasets. Therefore, studies have focused on overcoming this problem. In this study, the detection performance of five convolutional neural network (CNN) models (MobileNet V2 and NASNetMobile (lightweight models); VGG16, Inception V3, and ResNet (heavier weight models)) on medical images was compared to establish a classification model for epidemiological research. Brain scan image data were collected from 99 subjects, and 4711 images were used. Demographic data were compared using the chi-squared test and one-way analysis of variance with Bonferroni's post hoc test. Accuracy and loss functions were used to evaluate the performance of CNN models. The cognitive abilities screening instrument and mini mental state exam scores of subjects with a clinical dementia rating (CDR) of 2 were considerably lower than those of subjects with a CDR of 1 or 0.5. This study analyzed the classification performance of various CNN models for medical images and proved the effectiveness of transfer learning in identifying the mild cognitive impairment, mild AD, and moderate AD scoring based on SPECT images.

Partially Supervised Compatibility Modeling.

Fashion Compatibility Modeling (FCM), which aims to automatically evaluate whether a given set of fashion items makes a compatible outfit, has attracted increasing research attention. Recent studies have demonstrated the benefits of conducting the item representation disentanglement towards FCM. Although these efforts have achieved prominent progress, they still perform unsatisfactorily, as they mainly investigate the visual content of fashion items, while overlooking the semantic attributes of items (e.g., color and pattern), which could largely boost the model performance and interpretability. To address this issue, we propose to comprehensively explore the visual content and attributes of fashion items towards FCM. This problem is non-trivial considering the following challenges: a) how to utilize the irregular attribute labels of items to partially supervise the attribute-level representation learning of fashion items; b) how to ensure the intact disentanglement of attribute-level representations; and c) how to effectively sew the multiple granulairites (i.e, coarse-grained item-level and fine-grained attribute-level) information to enable performance improvement and interpretability. To address these challenges, in this work, we present a partially supervised outfit compatibility modeling scheme (PS-OCM). In particular, we first devise a partially supervised attribute-level embedding learning component to disentangle the fine-grained attribute embeddings from the entire visual feature of each item. We then introduce a disentangled completeness regularizer to prevent the information loss during disentanglement. Thereafter, we design a hierarchical graph convolutional network, which seamlessly integrates the attribute- and item-level compatibility modeling, and enables the explainable compatibility reasoning. Extensive experiments on the real-world dataset demonstrate that our PS-OCM significantly outperforms the state-of-the-art baselines. We have released our source codes and well-trained models to benefit other researchers (https://site2750.wixsite.com/ps-ocm).

Interactive Somatosensory Games in Rehabilitation Training for Older Adults With Mild Cognitive Impairment: Usability Study.

BACKGROUND: In aging societies, dementia risk increases with advancing age, increasing the incidence of dementia-related degenerative diseases and other complications, especially fall risk. Dementia also escalates the care burden, impacting patients, their families, social welfare institutions, and the social structure and medical system. OBJECTIVE: In elderly dementia, traditional card recognition rehabilitation (TCRR) does not effectively increase one's autonomy. Therefore, from the usability perspective, we used the Tetris game as a reference to develop an interactive somatosensory game rehabilitation (ISGR) with nostalgic style for elders with mild cognitive impairment (MCI). Through intuitive gesture-controlled interactive games, we evaluated subjective feelings concerning somatosensory game integration into rehabilitation to explore whether the ISGR could improve the willingness to use and motivation for rehabilitation among elders with MCI. METHODS: A total of 15 elders with MCI (7 males and 8 females with an average age of 78.4 years) underwent 2 experiments for 15 minutes. During experiment 1, TCRR was performed, followed by completing the questionnaire of the System Usability Scale (SUS). After 3-5 minutes, the second experiment (the ISGR) was conducted, followed by completing another SUS. We used SUS to explore differences in impacts of TCRR and ISGR on willingness to use among elders with MCI. In addition, we further investigated whether the factor of gender or prior rehabilitation experience would affect the rehabilitation willingness or not. RESULTS: The novel ISGR made the elderly feel interested and improved their willingness for continuous rehabilitation. According to the overall SUS score, the ISGR had better overall usability performance (73.7) than the TCRR (58.0) (t28=-4.62, P<.001). Furthermore, the ISGR individual item scores of "Willingness to Use" (t28=-8.27, P<.001), "Easy to Use" (t28=-3.17, P<.001), "System Integration" (t28=-5.07, P<.001), and "Easy to Learn" (t28=-2.81, P<.001) were better than TCRR. The somatosensory game was easier to learn and master for females than for males (t13=2.71, P=.02). Besides, the ISGR was easier to use (t12=-2.50, P=.02) and learn (t14=-3.33, P<.001) for those without prior rehabilitation experience. The result indicates that for elders with no rehabilitation experience ISGR was easier to use and simpler to learn than TCRR. CONCLUSIONS: Regardless of prior rehabilitation experience, the ISGR developed in this study was easy to learn and effective in continuously improving willingness to use. Furthermore, the adoption of a nostalgic game design style served the function of cognitive training and escalated interest in rehabilitation. The ISGR also improved user stickiness by introducing different game scenarios and difficulties, increasing long-term interest and motivation for rehabilitation. For future research on the adoption of interactive somatosensory games in rehabilitation, additional rehabilitation movements can be developed to benefit the elderly with MCI.

Significant symptoms and nonsymptom-related factors for malaria diagnosis in endemic regions of Indonesia.

OBJECTIVES: This study aims to identify significant symptoms and nonsymptom-related factors for malaria diagnosis in endemic regions of Indonesia. METHODS: Medical records are collected from patients suffering from malaria and other febrile diseases from public hospitals in endemic regions of Indonesia. Interviews with eight Indonesian medical doctors are conducted. Feature selection and machine learning techniques are used to develop malaria classifiers for identifying significant symptoms and nonsymptom-related factors. RESULTS: Seven significant symptoms (duration of fever, headache, nausea and vomiting, heartburn, severe symptom, dizziness, and joint pain) and patients' history of malaria as a nonsymptom-related factor contribute most to malaria diagnosis. As a symptom, fever duration is more significant than temperature or fever for distinguishing malaria from other febrile diseases. Shivering, fever, and sweating (known to indicate malaria presence in Indonesia) are shown to be less significant than other symptoms in endemic regions. CONCLUSIONS: Three most suitable malaria classifiers have been developed to identify the significant features that can be used to predict malaria as distinct from other febrile diseases. With extensive experiments on the classifiers, the significant features identified can help medical doctors in the clinical diagnosis of malaria and raise public awareness of significant malaria symptoms at early stages.

Semantics-Preserving Graph Propagation for Zero-Shot Object Detection.

Most existing object detection models are restricted to detecting objects from previously seen categories, an approach that tends to become infeasible for rare or novel concepts. Accordingly, in this paper, we explore object detection in the context of zero-shot learning, i.e., Zero-Shot Object Detection (ZSD), to concurrently recognize and localize objects from novel concepts. Existing ZSD algorithms are typically based on a simple mapping-transfer strategy that is susceptible to the domain shift problem. To resolve this problem, we propose a novel Semantics-Preserving Graph Propagation model for ZSD based on Graph Convolutional Networks (GCN). More specifically, we employ a graph construction module to flexibly build category graphs by incorporating diverse correlations between category nodes; this is followed by two semantics preserving modules that enhance both category and region representations through a multi-step graph propagation process. Compared to existing mapping-transfer based methods, both the semantic description and semantic structural knowledge exhibited in prior category graphs can be effectively leveraged to boost the generalization capability of the learned projection function via knowledge transfer, thereby providing a solution to the domain shift problem. Experiments on existing seen/unseen splits of three popular object detection datasets demonstrate that the proposed approach performs favorably against state-of-the-art ZSD methods.

Is the linear modeling technique good enough for optimal form design? A comparison of quantitative analysis models.

How to design highly reputable and hot-selling products is an essential issue in product design. Whether consumers choose a product depends largely on their perception of the product image. A consumer-oriented design approach presented in this paper helps product designers incorporate consumers' perceptions of product forms in the design process. The consumer-oriented design approach uses quantification theory type I, grey prediction (the linear modeling technique), and neural networks (the nonlinear modeling technique) to determine the optimal form combination of product design for matching a given product image. An experimental study based on the concept of Kansei Engineering is conducted to collect numerical data for examining the relationship between consumers' perception of product image and product form elements of personal digital assistants (PDAs). The result of performance comparison shows that the QTTI model is good enough to help product designers determine the optimal form combination of product design. Although the PDA form design is used as a case study, the approach is applicable to other consumer products with various design elements and product images. The approach provides an effective mechanism for facilitating the consumer-oriented product design process.

Human performance interfaces in air traffic control.

This paper examines how human performance factors in air traffic control (ATC) affect each other through their mutual interactions. The paper extends the conceptual SHEL model of ergonomics to describe the ATC system as human performance interfaces in which the air traffic controllers interact with other human performance factors including other controllers, software, hardware, environment, and organisation. New research hypotheses about the relationships between human performance interfaces of the system are developed and tested on data collected from air traffic controllers, using structural equation modelling. The research result suggests that organisation influences play a more significant role than individual differences or peer influences on how the controllers interact with the software, hardware, and environment of the ATC system. There are mutual influences between the controller-software, controller-hardware, controller-environment, and controller-organisation interfaces of the ATC system, with the exception of the controller-controller interface. Research findings of this study provide practical insights in managing human performance interfaces of the ATC system in the face of internal or external change, particularly in understanding its possible consequences in relation to the interactions between human performance factors.