Search

With Artificial Intelligence (AI) becoming a powerful tool for education (Zawacki-Richter et al., 2019), this chapter describes the concept of combining generative and traditional AI, citizen-science physiological, neuroergonomic wearables and environmental sensors into activities for learners to understand their own well-being and emotional states better with a view to developing empathy and environmental stewardship. Alongside bespoke and affordable wearables (DIY EEG headsets and biometric wristbands), interpretable AI and data science are used for learners to explore how the environment affects them physiologically and mentally in authentic environments. For example, relationships between environmental changes (e.g. poorer air quality) and their well-being (e.g. cognitive functioning) can be discovered. This is particularly crucial, as relevant knowledge can influence the way people treat the environment, as suggested by the disciplines of environmental neuroscience and environmental psychology (Doell et al., 2023). Yet, according to Palme and Salvati, there have been relatively few studies on the relationships between microclimates and human health and emotions (Palme and Salvati, 2021). As anthropogenic environmental pollution is becoming a prevalent problem, our research also aims to leverage on generative AI to introduce hypothetical scenarios of the environment as emotionally strong stimuli of relevance to the learners. This would provoke an emotional response for them to learn about their own physiological and neurological responses (using neuro-physiological data). Ultimately, we hope to establish a bidirectional understanding of how the environment affects humans physiologically and mentally; after which, to gain insights as to how AI can be used to effectively foster empathy, pro-environmental attitudes and stewardship., Comment: 14 pages, 8 figures

Nowadays, cardiac diagnosis largely depends on left ventricular function assessment. With the help of the segmentation deep learning model, the assessment of the left ventricle becomes more accessible and accurate. However, deep learning technique still faces two main obstacles: the difficulty in acquiring sufficient training data and time-consuming in developing quality models. In the ordinary data acquisition process, the dataset was selected randomly from a large pool of unlabeled images for labeling, leading to massive labor time to annotate those images. Besides that, hand-designed model development is strenuous and also costly. This paper introduces a pipeline that relies on Active Learning to ease the labeling work and utilizes Neural Architecture Search's idea to design the adequate deep learning model automatically. We called this Fully automated machine learning pipeline for echocardiogram segmentation. The experiment results show that our method obtained the same IOU accuracy with only two-fifths of the original training dataset, and the searched model got the same accuracy as the hand-designed model given the same training dataset.