Using Graphs to Perform Effective Sensor-Based Human Activity Recognition in Smart Homes.

There has been a resurgence of applications focused on human activity recognition (HAR) in smart homes, especially in the field of ambient intelligence and assisted-living technologies. However, such applications present numerous significant challenges to any automated analysis system operating in the real world, such as variability, sparsity, and noise in sensor measurements. Although state-of-the-art HAR systems have made considerable strides in addressing some of these challenges, they suffer from a practical limitation: they require successful pre-segmentation of continuous sensor data streams prior to automated recognition, i.e., they assume that an oracle is present during deployment, and that it is capable of identifying time windows of interest across discrete sensor events. To overcome this limitation, we propose a novel graph-guided neural network approach that performs activity recognition by learning explicit co-firing relationships between sensors. We accomplish this by learning a more expressive graph structure representing the sensor network in a smart home in a data-driven manner. Our approach maps discrete input sensor measurements to a feature space through the application of attention mechanisms and hierarchical pooling of node embeddings. We demonstrate the effectiveness of our proposed approach by conducting several experiments on CASAS datasets, showing that the resulting graph-guided neural network outperforms the state-of-the-art method for HAR in smart homes across multiple datasets and by large margins. These results are promising because they push HAR for smart homes closer to real-world applications.

Towards Learning Discrete Representations via Self-Supervision for Wearables-Based Human Activity Recognition.

Human activity recognition (HAR) in wearable and ubiquitous computing typically involves translating sensor readings into feature representations, either derived through dedicated pre-processing procedures or integrated into end-to-end learning approaches. Independent of their origin, for the vast majority of contemporary HAR methods and applications, those feature representations are typically continuous in nature. That has not always been the case. In the early days of HAR, discretization approaches had been explored-primarily motivated by the desire to minimize computational requirements on HAR, but also with a view on applications beyond mere activity classification, such as, for example, activity discovery, fingerprinting, or large-scale search. Those traditional discretization approaches, however, suffer from substantial loss in precision and resolution in the resulting data representations with detrimental effects on downstream analysis tasks. Times have changed, and in this paper, we propose a return to discretized representations. We adopt and apply recent advancements in vector quantization (VQ) to wearables applications, which enables us to directly learn a mapping between short spans of sensor data and a codebook of vectors, where the index comprises the discrete representation, resulting in recognition performance that is at least on par with their contemporary, continuous counterparts-often surpassing them. Therefore, this work presents a proof of concept for demonstrating how effective discrete representations can be derived, enabling applications beyond mere activity classification but also opening up the field to advanced tools for the analysis of symbolic sequences, as they are known, for example, from domains such as natural language processing. Based on an extensive experimental evaluation of a suite of wearable-based benchmark HAR tasks, we demonstrate the potential of our learned discretization scheme and discuss how discretized sensor data analysis can lead to substantial changes in HAR.

The Lifespan of Human Activity Recognition Systems for Smart Homes.

With the growing interest in smart home environments and in providing seamless interactions with various smart devices, robust and reliable human activity recognition (HAR) systems are becoming essential. Such systems provide automated assistance to residents or to longitudinally monitor their daily activities for health and well-being assessments, as well as for tracking (long-term) behavior changes. These systems thus contribute towards an understanding of the health and continued well-being of residents. Smart homes are personalized settings where residents engage in everyday activities in their very own idiosyncratic ways. In order to provide a fully functional HAR system that requires minimal supervision, we provide a systematic analysis and a technical definition of the lifespan of activity recognition systems for smart homes. Such a designed lifespan provides for the different phases of building the HAR system, where these different phases are motivated by an application scenario that is typically observed in the home setting. Through the aforementioned phases, we detail the technical solutions that are required to be developed for each phase such that it becomes possible to derive and continuously improve the HAR system through data-driven procedures. The detailed lifespan can be used as a framework for the design of state-of-the-art procedures corresponding to the different phases.

Complex Deep Neural Networks from Large Scale Virtual IMU Data for Effective Human Activity Recognition Using Wearables.

Supervised training of human activity recognition (HAR) systems based on body-worn inertial measurement units (IMUs) is often constrained by the typically rather small amounts of labeled sample data. Systems like IMUTube have been introduced that employ cross-modality transfer approaches to convert videos of activities of interest into virtual IMU data. We demonstrate for the first time how such large-scale virtual IMU datasets can be used to train HAR systems that are substantially more complex than the state-of-the-art. Complexity is thereby represented by the number of model parameters that can be trained robustly. Our models contain components that are dedicated to capture the essentials of IMU data as they are of relevance for activity recognition, which increased the number of trainable parameters by a factor of 1100 compared to state-of-the-art model architectures. We evaluate the new model architecture on the challenging task of analyzing free-weight gym exercises, specifically on classifying 13 dumbbell execises. We have collected around 41 h of virtual IMU data using IMUTube from exercise videos available from YouTube. The proposed model is trained with the large amount of virtual IMU data and calibrated with a mere 36 min of real IMU data. The trained model was evaluated on a real IMU dataset and we demonstrate the substantial performance improvements of 20% absolute F1 score compared to the state-of-the-art convolutional models in HAR.

Sphingosine-1 Phosphate Lyase Regulates Sensitivity of Pancreatic Beta-Cells to Lipotoxicity.

Elevated levels of free fatty acids (FFAs) have been related to pancreatic beta-cell failure in type 2 diabetes (T2DM), though the underlying mechanisms are not yet fully understood. FFAs have been shown to dysregulate formation of bioactive sphingolipids, such as ceramides and sphingosine-1 phosphate (S1P) in beta-cells. The aim of this study was to analyze the role of sphingosine-1 phosphate lyase (SPL), a key enzyme of the sphingolipid pathway that catalyzes an irreversible degradation of S1P, in the sensitivity of beta-cells to lipotoxicity. To validate the role of SPL in lipotoxicity, we modulated SPL expression in rat INS1E cells and in human EndoC-ßH1 beta-cells. SPL overexpression in INS1E cells (INS1E-SPL), which are characterized by a moderate basal expression level of SPL, resulted in an acceleration of palmitate-mediated cell viability loss, proliferation inhibition and induction of oxidative stress. SPL overexpression affected the mRNA expression of ER stress markers and mitochondrial chaperones. In contrast to control cells, in INS1E-SPL cells no protective effect of oleate was detected. Moreover, Plin2 expression and lipid droplet formation were strongly reduced in OA-treated INS1E-SPL cells. Silencing of SPL in human EndoC-ßH1 beta-cells, which are characterized by a significantly higher SPL expression as compared to rodent beta-cells, resulted in prevention of FFA-mediated caspase-3/7 activation. Our findings indicate that an adequate control of S1P degradation by SPL might be crucially involved in the susceptibility of pancreatic beta-cells to lipotoxicity.

A Combined Deep Learning GRU-Autoencoder for the Early Detection of Respiratory Disease in Pigs Using Multiple Environmental Sensors.

We designed and evaluated an assumption-free, deep learning-based methodology for animal health monitoring, specifically for the early detection of respiratory disease in growing pigs based on environmental sensor data. Two recurrent neural networks (RNNs), each comprising gated recurrent units (GRUs), were used to create an autoencoder (GRU-AE) into which environmental data, collected from a variety of sensors, was processed to detect anomalies. An autoencoder is a type of network trained to reconstruct the patterns it is fed as input. By training the GRU-AE using environmental data that did not lead to an occurrence of respiratory disease, data that did not fit the pattern of "healthy environmental data" had a greater reconstruction error. All reconstruction errors were labelled as either normal or anomalous using threshold-based anomaly detection optimised with particle swarm optimisation (PSO), from which alerts are raised. The results from the GRU-AE method outperformed state-of-the-art techniques, raising alerts when such predictions deviated from the actual observations. The results show that a change in the environment can result in occurrences of pigs showing symptoms of respiratory disease within 1⁻7 days, meaning that there is a period of time during which their keepers can act to mitigate the negative effect of respiratory diseases, such as porcine reproductive and respiratory syndrome (PRRS), a common and destructive disease endemic in pigs.

Multi-part segmentation for porcine offal inspection with auto-context and adaptive atlases.

Extensions to auto-context segmentation are proposed and applied to segmentation of multiple organs in porcine offal as a component of an envisaged system for post-mortem inspection at abbatoir. In common with multi-part segmentation of many biological objects, challenges include variations in configuration, orientation, shape, and appearance, as well as inter-part occlusion and missing parts. Auto-context uses context information about inferred class labels and can be effective in such settings. Whereas auto-context uses a fixed prior atlas, we describe an adaptive atlas method better suited to represent the multimodal distribution of segmentation maps. We also design integral context features to enhance context representation. These methods are evaluated on a dataset captured at abbatoir and compared to a method based on conditional random fields. Results demonstrate the appropriateness of auto-context and the beneficial effects of the proposed extensions for this application.

The genetic basis of novel water utilisation and drinking behaviour traits and their relationship with biological performance in turkeys.

BACKGROUND: There is increasing interest in the definition, measurement and use of traits associated with water use and drinking behaviour, mainly because water is a finite resource and its intake is an important part of animal health and well-being. Analysis of such traits has received little attention, due in part to the lack of appropriate technology to measure drinking behaviour. We exploited novel equipment to collect water intake data in two lines of turkey (A: 27,415 and B: 12,956 birds). The equipment allowed continuous recording of individual visits to the water station in a group environment. Our aim was to identify drinking behaviour traits of biological relevance, to estimate their genetic parameters and their genetic relationships with performance traits, and to identify drinking behaviour strategies among individuals. RESULTS: Visits to the drinkers were clustered into bouts, i.e. time intervals spent in drinking-related activity. Based on this, biologically relevant traits were defined: (1) number of visits per bout, (2) water intake per bout, (3) drinking time per bout, (4) drinking rate, (5) daily bout frequency, (6) daily bout duration, (7) daily drinking time and (8) daily water intake. Heritability estimates for most drinking behaviour traits were moderate to high and the most highly heritable traits were drinking rate (0.49 and 0.50) and daily drinking time (0.35 and 0.46 in lines A and B, respectively). Genetic correlations between drinking behaviour and performance traits were low except for moderate correlations between daily water intake and weight gain (0.46 and 0.47 in lines A and B, respectively). High estimates of breeding values for weight gain were found across the whole range of estimated breeding values for daily water intake, daily drinking time and water intake per bout. CONCLUSIONS: We show for the first time that drinking behaviour traits are moderately to highly heritable. Low genetic and phenotypic correlations with performance traits suggest that current breeding goals have not and will not affect normal water drinking behaviour. Birds express a wide range of different drinking behaviour strategies, which can be suitable to a wide range of environments and production systems.

Sensitivity profile of the human EndoC-ßH1 beta cell line to proinflammatory cytokines.

AIMS/HYPOTHESIS: The aim of this study was to perform a detailed analysis of cytokine toxicity in the new human EndoC-ßH1 beta cell line. METHODS: The expression profile of the antioxidative enzymes in the new human EndoC-ßH1 beta cells was characterised and compared with that of primary beta cells in the human pancreas. The effects of proinflammatory cytokines on reactive oxygen species formation, insulin secretory responsiveness and apoptosis of EndoC-ßH1 beta cells were determined. RESULTS: EndoC-ßH1 beta cells were sensitive to the toxic action of proinflammatory cytokines. Glucose-dependent stimulation of insulin secretion and an increase in the ATP/ADP ratio was abolished by proinflammatory cytokines without induction of IL-1ß expression. Cytokine-mediated caspase-3 activation was accompanied by reactive oxygen species formation and developed more slowly than in rodent beta cells. Cytokines transiently increased the expression of unfolded protein response genes, without inducing endoplasmic reticulum stress-marker genes. Cytokine-mediated NFκB activation was too weak to induce inducible nitric oxide synthase expression. The resultant lack of nitric oxide generation in EndoC-ßH1 cells, in contrast to rodent beta cells, makes these cells dependent on exogenously generated nitric oxide, which is released from infiltrating immune cells in human type 1 diabetes, for full expression of proinflammatory cytokine toxicity. CONCLUSIONS/INTERPRETATION: EndoC-ßH1 beta cells are characterised by an imbalance between H2O2-generating and -inactivating enzymes, and react to cytokine exposure in a similar manner to primary human beta cells. They are a suitable beta cell surrogate for cytokine-toxicity studies.

Porcine lie detectors: Automatic quantification of posture state and transitions in sows using inertial sensors.

This paper presents a novel approach to automated classification and quantification of sow postures and posture transitions that may enable large scale and accurate continuous behaviour assessment on farm. Automatic classification and quantification of postures and posture transitions in domestic animals has substantial potential to enhance their welfare and productivity. Analysis of such behaviours in farrowing sows can highlight the need for human intervention or lead to the prediction of movement patterns that are potentially dangerous for their piglets, such as crushing when the sow lies down. Data were recorded by a tri-axial accelerometer secured to the hind-end of each sow, in a deployment that involved six sows over the period around parturition. The posture state (standing, sitting, lateral and sternal lying) was automatically classified for the full dataset with a mean F1 score (a measure of predictive performance between 0 and 1) of 0.78. Sitting was shown to present a greater challenge to classification with a F1 score of 0.54, compared to the lateral lying postures, which were classified with an average F1 score of 0.91. Posture transitions were detected with a F1 score of 0.79. We automatically extracted and visualized a range of features that characterise the manner in which the sows changed posture in order to provide comparative descriptors of sow activity and lying style that can be used to assess the influence of genetics or housing design. The methodology presented in this paper can be applied in large scale deployments with potential for enhancing animal welfare and productivity on farm.

Antagonism Between Saturated and Unsaturated Fatty Acids in ROS Mediated Lipotoxicity in Rat Insulin-Producing Cells.

BACKGROUND/AIMS: Elevated levels of non-esterified fatty acids (NEFAs) are under suspicion to mediate ß-cell dysfunction and ß-cell loss in type 2 diabetes, a phenomenon known as lipotoxicity. Whereas saturated fatty acids show a strong cytotoxic effect upon insulin-producing cells, unsaturated fatty acids are not toxic and can even prevent toxicity. Experimental evidence suggests that oxidative stress mediates lipotoxicity and there is evidence that the subcellular site of ROS formation is the peroxisome. However, the interaction between unsaturated and saturated NEFAs in this process is unclear. METHODS: Toxicity of rat insulin-producing cells after NEFA incubation was measured by MTT and caspase assays. NEFA induced H2O2 formation was quantified by organelle specific expression of the H2O2 specific fluorescence sensor protein HyPer. RESULTS: The saturated NEFA palmitic acid had a significant toxic effect on the viability of rat insulin-producing cells. Unsaturated NEFAs with carbon chain lengths >14 showed, irrespective of the number of double bonds, a pronounced protection against palmitic acid induced toxicity. Palmitic acid induced H2O2 formation in the peroxisomes of insulin-producing cells. Oleic acid incubation led to lipid droplet formation, but in contrast to palmitic acid induced neither an ER stress response nor peroxisomal H2O2 generation. Furthermore, oleic acid prevented palmitic acid induced H2O2 production in the peroxisomes. CONCLUSION: Thus unsaturated NEFAs prevent deleterious hydrogen peroxide generation during peroxisomal ß-oxidation of long-chain saturated NEFAs in rat insulin-producing cells.

Mechanisms of lipotoxicity-induced dysfunction and death of human pancreatic beta cells under obesity and type 2 diabetes conditions.

The term "pancreatic beta-cell lipotoxicity" refers to the detrimental effects of free fatty acids (FFAs) on a wide variety of cellular functions. Basic research in the field has primarily analyzed the effects of palmitic acid and oleic acid. The focus on these two physiological FFAs, however, ignores differences in chain length and degree of saturation. In order to gain a comprehensive understanding of the lipotoxic mechanisms, a wide range of structurally related FFAs should be investigated. Structure-activity relationship analyses of FFAs in the human EndoC-ßH1 beta-cell line have provided a deep insight into the mechanisms of beta-cell lipotoxicity. This review focuses on the effects of a wide range of FFAs with crucial structural determinants for the development of lipotoxicity in human beta cells and documents an association between increased triglyceride stores in obesity and in type 2 diabetes.

Oxidative and ER stress by elevated insulin biosynthesis and palmitic acid in insulin-producing cells.

The early phase of type 2 diabetes mellitus (T2DM) is characterised by insulin resistance, which can initially be compensated by elevated insulin secretion. However, as postulated by the workload hypothesis, over time harming insulin requirements contribute to ß-cell dysfunction and death. The mechanisms behind this transition are complex and not fully understood but involve factors such as endoplasmic reticulum (ER) stress raised by gluco/lipotoxicity. To investigate the effect of excessive insulin folding on ER luminal H2O2 generation, ER stress and viability, insulin was expressed glucose-independently by a doxycycline-regulated Tet-On system in insulin-producing RINm5F cells. Additionally, the effect of palmitic acid (PA) as a subsidiary T2DM-associated factor was examined in this model system. Elevated insulin expression increased ER luminal H2O2 concentration quantified by the fluorescent sensor protein TriPer and reduced viability, but did not activate apoptosis. However, when combined with PA, insulin expression resulted in a significant increase in ER stress and apoptosis. Expression of ER-localised catalase verified the specificity of the applied H2O2 detection method without attenuating ER stress, caspase activation or viability loss. These findings suggest that hyperinsulinism alone can cause increased ER luminal H2O2 generation, mild ER stress and reduced viability, while hyperinsulinism in combination with PA accelerates these processes and triggers apoptosis. The inability of ER catalase to counteract these effects suggests that further damaging factors besides H2O2 are involved in cell dysfunction. Finally, reducing the high insulin demand in the initial phase of T2DM may be crucial in preventing further ß-cell damage caused by gluco/lipotoxicity.

Potentiation of Lipotoxicity in Human EndoC-ßH1 ß-Cells by Glucose is Dependent on the Structure of Free Fatty Acids.

SCOPE: Lipotoxicity is a significant element in the development of type 2 diabetes mellitus (T2DM). Since pro-diabetic nutritional patterns are associated with hyperglycemia as well as hyperlipidemia, the study analyzes the effects of combining these lipid and carbohydrate components with a special focus on the structural fatty acid properties such as increasing chain length (C16-C20) and degree of saturation with regard to the role of glucolipotoxicity in human EndoC-ßH1 ß-cells. METHODS AND RESULTS: ß-cell death induced by saturated FFAs is potentiated by high concentrations of glucose in a chain length-dependent manner starting with stearic acid (C18:0), whereas toxicity remains unchanged in the case of monounsaturated FFAs. Interference with FFA desaturation by overexpression and inhibition of stearoyl-CoA-desaturase, which catalyzes the rate-limiting step in the conversion of long-chain saturated into corresponding monounsaturated FFAs, does not affect the potentiating effect of glucose, but FFA desaturation reduces lipotoxicity and plays an important role in the formation of lipid droplets. Crucial elements underlying glucolipotoxicity are ER stress induction and cardiolipin peroxidation in the mitochondria. CONCLUSION: In the context of nutrition, the data emphasize the importance of the lipid component in glucolipotoxicity related to the development of ß-cell dysfunction and death in the manifestation of T2DM.

Differential effects of saturated and unsaturated free fatty acids on ferroptosis in rat ß-cells.

Elevated plasma concentrations of saturated free fatty acids (SFAs) are involved in pancreatic ß-cell dysfunction and apoptosis, referred to as lipotoxicity. However, in contrast to apoptosis, the involvement of ferroptosis, as a distinct type of oxidative regulated cell death in ß-cell lipotoxicity remains elusive. Therefore, the aim of this study was to determine the effects of various free fatty acids on ferroptosis induction in rat insulin-producing ß-cells. Herein, rat insulin-producing ß-cells underwent lipid peroxidation in the presence of long-chain SFAs and ω-6-polyunsaturated fatty acids (PUFAs), but only the latter induced ferroptosis. On the other hand, the ω-3-PUFA α-linolenate did not induce ferroptosis but sensitized insulin-producing ß-cells to SFA-mediated lipid peroxidation. While the monounsaturated fatty acid oleate, overexpression of glutathione peroxidase 4, and the specific ferroptosis inhibitor ferrostatin-1 significantly abrogated lipid peroxidation, neither glutathione peroxidase 4 nor ferrostatin-1 affected palmitate-mediated toxicity. Site-specific expression of catalase in cytosol, mitochondria, and ER attenuated lipid peroxidation, indicating the contribution of metabolically generated H2O2 from all three subcellular compartments. These observations suggest that only ω-6-PUFAs reach the thresholds of lipid peroxidation required for ferroptosis, whereas SFAs favour apoptosis in ß-cells. Hence, avoiding an excessive dietary intake of ω-6-PUFAs might be a crucial prerequisite for prevention of reactive oxygen species-mediated ferroptosis in insulin-producing cells.

The central role of glutathione peroxidase 4 in the regulation of ferroptosis and its implications for pro-inflammatory cytokine-mediated beta-cell death.

Pro-inflammatory cytokines are crucial mediators of beta-cell destruction in type 1 diabetes mellitus (T1DM). The involvement of ferroptosis as a form of oxidative non-apoptotic cell death in T1DM pathogenesis has not been elucidated so far. Moreover, the role of glutathione peroxidase 4 (GPx4) as an antioxidative enzyme and a major regulator of ferroptosis remains elusive. Assessment of GPx4 expression in different pancreatic islet cell types revealed a predominant expression in beta-cells. Silencing of GPx4 by RNA interference and exposure to tert-butyl hydroperoxide (tert-BHP) caused ferroptosis in rat pancreatic beta-cells as evidenced by non-apoptotic cell death in association with increased lipid peroxidation, disturbed ATP synthesis, reduced GSH content, and GPx4 degradation. GPx4 overexpression as well as the ferroptosis inhibitor ferrostatin-1 effectively attenuated beta-cell death induced by tert-BHP. Notably, beta-cell toxic cytokines did not induce ferroptosis although beta-cells underwent cell death. Inhibition of iNOS by Nω-nitro-L-arginine however led to a massive lipid peroxidation upon exposure to pro-inflammatory cytokines. Hence, nitric oxide produced during pro-inflammatory cytokine action prevents the induction of ferroptosis, thereby favouring apoptosis as a primary cell death mechanism. The extraordinarily high abundance of the phospholipid hydroperoxidase GPx4 in beta-cells in contrast to the very low expression in other islet cell types points to a susceptibility of beta-cells to the accumulation of toxic lipid peroxides. Overall, these data strongly suggest that GPx4 is indispensable for beta-cell function under physiological conditions. On the other hand, our results exclude an involvement of ferroptosis as an alternative beta-cell death mode under pro-inflammatory cytokine attack.

Estimation of Instantaneous Oxygen Uptake During Exercise and Daily Activities Using a Wearable Cardio-Electromechanical and Environmental Sensor.

OBJECTIVE: To estimate instantaneous oxygen uptake VO2 with a small, low-cost wearable sensor during exercise and daily activities in order to enable monitoring of energy expenditure (EE) in uncontrolled settings. We aim to do so using a combination of seismocardiogram (SCG), electrocardiogram (ECG) and atmospheric pressure (AP) signals obtained from a minimally obtrusive wearable device. METHODS: In this study, subjects performed a treadmill protocol in a controlled environment and an outside walking protocol in an uncontrolled environment. During testing, the COSMED K5 metabolic system collected gold standard breath-by-breath (BxB) data and a custom-built wearable patch placed on the mid-sternum collected SCG, ECG and AP signals. We extracted features from these signals to estimate the BxB VO2 data obtained from the COSMED system. RESULTS: In estimating instantaneous VO2, we achieved our best results on the treadmill protocol using a combination of SCG (frequency) and AP features (RMSE of 3.68 ± 0.98 ml/kg/min and R2 of 0.77). For the outside protocol, we achieved our best results using a combination of SCG (frequency), ECG and AP features (RMSE of 4.3 ± 1.47 ml/kg/min and R2 of 0.64). In estimating VO2 consumed over one minute intervals during the protocols, our median percentage error was 15.8[Formula: see text] for the treadmill protocol and 20.5[Formula: see text] for the outside protocol. CONCLUSION: SCG, ECG and AP signals from a small wearable patch can enable accurate estimation of instantaneous VO2 in both controlled and uncontrolled settings. SCG signals capturing variation in cardio-mechanical processes, AP signals, and state of the art machine learning models contribute significantly to the accurate estimation of instantaneous VO2. SIGNIFICANCE: Accurate estimation of VO2 with a low cost, minimally obtrusive wearable patch can enable the monitoring of VO2 and EE in everyday settings and make the many applications of these measurements more accessible to the general public.

Toxicity of fatty acid profiles of popular edible oils in human EndoC-ßH1 beta-cells.

An inappropriate diet, particularly excessive consumption of dietary fats and oils, may have a major negative impact on beta-cell function and cause type 2 diabetes mellitus. To investigate this issue, we examined the toxicity of free fatty acid (FFA) compositions mirroring the FFA profiles of various popular edible oils in human EndoC-ßH1 beta-cells and in rat islets. For this purpose, we made compositions consisting exclusively of various FFAs in different volumetric percentages mimicking these oils and additionally mixtures of these compositions. Human EndoC-ßH1 beta-cells were incubated with different oil compositions and the toxicity, lipid droplet formation, ER-stress, and H2O2 production were analyzed. Compositions with prominent content of saturated as well as unsaturated long-chain FFAs showed moderate but significant toxicity both in human EndoC-ßH1 beta-cells and rat islets, however, without further measurable metabolic impairments. On the other hand compositions with high content of medium-chain FFAs revealed no toxicity. A composition with 50% of the very long-chain unsaturated FFA erucic acid caused high toxicity with concomitant peroxisomal H2O2 production. The toxicity of FFAs to human EndoC-ßH1 beta-cells was dampened in mixtures of FFA compositions with a significant content of medium-chain FFAs, but not with a significant proportion of unsaturated FFAs.

Measuring Self-Esteem with Passive Sensing.

Self-esteem encompasses how individuals evaluate themselves and is an important contributor to their success. Self-esteem has been traditionally measured using survey-based methodologies. However, surveys suffer from limitations such as retrospective recall and reporting biases, leading to a need for proactive measurement approaches. Our work uses smartphone sensors to predict self-esteem and is situated in a multimodal sensing study on college students for five weeks. We use theory-driven features, such as phone communications and physical activity to predict three dimensions, performance, social, and appearance self-esteem. We conduct statistical modeling including linear, ensemble, and neural network regression to measure self-esteem. Our best model predicts self-esteem with a high correlation (r) of 0.60 and low SMAPE of 7.26% indicating high predictive accuracy. We inspect the top features finding theoretical alignment; for example, social interaction significantly contributes to performance and appearance-based self-esteem, whereas, and physical activity is the most significant contributor towards social self-esteem. Our work reveals the efficacy of passive sensors for predicting self-esteem, and we situate our observations with literature and discuss the implications of our work for tailored interventions and improving wellbeing.

Automated Detection of Infant Holding Using Wearable Sensing: Implications for Developmental Science And Intervention.

Physical contact is critical for children's physical and emotional growth and well-being. Previous studies of physical contact are limited to relatively short periods of direct observation and self-report methods. These methods limit researchers' understanding of the natural variation in physical contact across families, and its specific impacts on child development. In this study we develop a mobile sensing platform that can provide objective, unobtrusive, and continuous measurements of physical contact in naturalistic home interactions. Using commercially available motion detectors, our model reaches an accuracy of 0.870 (std: 0.059) for a second-by-second binary classification of holding. In addition, we detail five assessment scenarios applicable to the development of activity recognition models for social science research, where required accuracy may vary as a function of the intended use. Finally, we propose a grand vision for leveraging mobile sensors to access high-density markers of multiple determinants of early parent-child interactions, with implications for basic science and intervention.