Ensemble Image Explainable AI (XAI) Algorithm for Severe Community-Acquired Pneumonia and COVID-19 Respiratory Infections

Since the onset of the COVID-19 pandemic in 2019, many clinical prognostic scoring tools have been proposed or developed to aid clinicians in the disposition and severity assessment of pneumonia. However, there is limited work that focuses on explaining techniques that are best suited for clinicians in their decision making. In this article, we present a new image explainability method named ensemble AI explainability (XAI), which is based on the SHAP and Grad-CAM++ methods. It provides a visual explanation for a deep learning prognostic model that predicts the mortality risk of community-acquired pneumonia and COVID-19 respiratory infected patients. In addition, we surveyed the existing literature and compiled prevailing quantitative and qualitative metrics to systematically review the efficacy of ensemble XAI, and to make comparisons with several state-of-the-art explainability methods (LIME, SHAP, saliency map, Grad-CAM, Grad-CAM++). Our quantitative experimental results have shown that ensemble XAI has a comparable absence impact (decision impact: 0.72, confident impact: 0.24). Our qualitative experiment, in which a panel of three radiologists were involved to evaluate the degree of concordance and trust in the algorithms, has showed that ensemble XAI has localization effectiveness (mean set accordance precision: 0.52, mean set accordance recall: 0.57, mean set F1: 0.50, mean set IOU: 0.36) and is the most trusted method by the panel of radiologists (mean vote: 70.2%). Finally, the deep learning interpretation dashboard used for the radiologist panel voting will be made available to the community. Our code is available at https://github.com/IHIS-HealthInsights/Interpretation-Methods-Voting-dashboard. © 2020 IEEE.

Utilizing a Telegram Quiz Bot to Promote Retrieval Practice

This study aims to support learning by encouraging the application of retrieval practice by students. This usage of retrieval practice aims to tackle pedagogical inadequacies such as social isolation and poor peer-to-peer communication. In this study, retrieval practice is incorporated into social media usage via a Telegram Quiz Bot. In addition to gamifying a student's learning experience, the Quiz Bot incorporated relevant topics taught in an environmental chemistry module. Students were required to compete in groups for accuracy and speed in answering questions. To evaluate the effectiveness of this Telegram quiz bot, a survey was conducted with 29 students enrolled in an environmental chemistry module at the National University of Singapore. Overall, positive responses were obtained, and students would recommend its incorporation in other modules as well. © 2021 IEEE.

Supporting Social and Learning Presence in the Revised Community of Inquiry Framework for Hybrid Learning

Teaching and learning became more challenging during the COVID-19 pandemic as classes moved online for both remote and hybrid learning. For hybrid learning, instructors harnessed various technologies to facilitate student-teacher engagement. Here we explored the use of an online collaborative platform, Miro Board, to aid teaching of organic chemistry for hybrid learning. We applied the revised Community of Inquiry (CoI) framework in our teaching exploration to support student- student and student-teacher interactions. Our experiences indicated that the use of Miro Board raises the social and learning presence in the revised CoI.

Social Frailty and Executive Function: Association with Geriatric Syndromes, Life Space and Quality of Life in Healthy Community-Dwelling Older Adults.

BACKGROUND: Despite emerging evidence about the association between social frailty and cognitive impairment, little is known about the role of executive function in this interplay, and whether the co-existence of social frailty and cognitive impairment predisposes to adverse health outcomes in healthy community-dwelling older adults. OBJECTIVES: We aim to examine independent associations between social frailty with the MMSE and FAB, and to determine if having both social frailty and cognitive impairment is associated with worse health outcomes than either or neither condition. METHODS: We studied 229 cognitively intact and functionally independent community-dwelling older adults (mean age= 67.2±7.43). Outcome measures comprise physical activity; physical performance and frailty; geriatric syndromes; life space and quality of life. We compared Chinese Mini Mental State Examination (CMMSE) and Chinese Frontal Assessment Battery (FAB) scores across the socially non-frail, socially pre-frail and socially frail. Participants were further recategorized into three subgroups (neither, either or both) based on presence of social frailty and cognitive impairment. Cognitive impairment was defined as a score below the educational adjusted cut-offs in either CMMSE or FAB. We performed logistic regression adjusted for significant covariates and mood to examine association with outcomes across the three subgroups. RESULTS: Compared with CMMSE, Chinese FAB scores significantly decreased across the social frailty spectrum (p<0.001), suggesting strong association between executive function with social frailty. We derived three subgroups relative to relationship with socially frailty and executive dysfunction: (i) Neither, N=140(61.1%), (ii) Either, N=79(34.5%), and (iii) Both, N=10(4.4%). Compared with neither or either subgroups, having both social frailty and executive dysfunction was associated with anorexia (OR=4.79, 95% CI= 1.04-22.02), near falls and falls (OR= 5.23, 95% CI= 1.10-24.90), lower life-space mobility (odds ratio, OR=9.80, 95% CI=2.07-46.31) and poorer quality of life (OR= 13.2, 95% CI= 2.38-73.4). CONCLUSION: Our results explicated the association of executive dysfunction with social frailty, and their synergistic relationship independent of mood with geriatric syndromes, decreased life space and poorer quality of life. In light of the current COVID-19 pandemic, the association between social frailty and executive dysfunction merits further study as a possible target for early intervention in relatively healthy older adults.

Coding a Telegram Quiz Bot to Aid Learners in Environmental Chemistry

In light of the COVID-19 pandemic, the transition from traditional face-to-face lectures to online learning, as well as the adoption of technology-enabled active learning in the classroom, has been accelerated. Here, we introduce the use of Telegram as a technological tool to support remote learning by helping students identify gaps in their content knowledge with ease and convenience. This was done by designing a quiz based on relevant topics on the Telegram Quiz Bot, while adding an element of gamification, whereby students compete in group chats for correctness and speed. Overall, the majority of the students responded positively to the use of the Telegram Quiz Bot. © Published 2021 by American Chemical Society and Division of Chemical Education, Inc.