Search

• Electrocardiogram (ECG) signals can be a potential alternative to continuous glucose monitoring (CGM) devices for detecting glycemic excursions. This includes hypoglycemia as well as hyperglycemia. • A combination of intra-beat (morphological features) and inter-beat (heart-rate variability features) information improves prediction performance compared to their independent use. • Changes in morphological patterns occur at varying glucose levels for different patients and at different times. • A personalized fusion-based approach combining morphological patterns at varying glucose levels leads to improved prediction performance over baseline models (trained only at conventional excursions thresholds). Monitoring glucose levels is critical for effective diabetes management. Continuous glucose monitoring devices estimate interstitial glucose levels and provide alerts for glycemic excursions. However, they are expensive and invasive. Therefore, low-cost, noninvasive alternatives are useful for patients with diabetes. In this article, we explore electrocardiogram signals as a potential alternative to detecting glycemic excursions by extracting intra-beat (beat-morphology) and inter-beat (heart rate variability) information. Unlike prior methods that focused only on the standard clinical excursion thresholds (70 mg/dL for hypoglycemia, 180 mg/dL for hyperglycemia), our proposed approach trains independent machine learning models at various excursion thresholds, aggregating their outputs for a final prediction. This allows learning morphological patterns in the neighborhood of the standard excursion thresholds. Our personalized fusion models achieve an AUC of 75 % for hypoglycemia and 78 % for hyperglycemia detection across patients, resulting in an average improvement of 4 % compared to the baseline models (trained using only standard clinical thresholds) for detecting glycemic excursions. We also find that combining morphology and HRV information outperforms using them individually (5 % for hypoglycemia and 6 % for hyperglycemia). The data used in this article was collected from 12 patients with type-1 diabetes, each monitored over a 14-day period at Texas Children's Hospital, Houston. The results indicate that a combination of morphological and HRV features is essential for noninvasive detection of glycemic excursions. Also, morphological changes can happen at varying glucose levels for different patients and capturing these changes provide valuable information that leads to improved prediction performance for detecting glycemic excursions. [ABSTRACT FROM AUTHOR]

• A home study of type 1 diabetes adolescents in a free-living condition is conducted to collect tremor and blood glucose data. • Four machine learning algorithms (decision tree, support vector machine, k-nearest neighbor, and bagged trees ensemble classifier) are applied to find the optimal model that detects hypoglycemia using tremor. • The study demonstrates the feasibility of detecting and predicting the onset of hypoglycemia based on hand tremor. Diabetes is a chronic disease where blood glucose (BG) concentrations are consistently high. A dangerous diabetes-associated condition is hypoglycemia, where BG level drops below the normal threshold. Hypoglycemia is often accommodated by tremors, sweating, fatigue, anxiety, lightheadedness, disorientation, irritability, and tachycardia. Very few studies are focused on detecting enhanced tremor as a peripheral physiological response to declining BG concentrations. This study undertakes a machine-learning approach to predict hypoglycemia using hand tremors. Tremors were detected and characterized by frequency and amplitude in non-hypoglycemic and hypoglycemic conditions. Accelerometers and continuous glucose monitoring devices recorded the tremor and BG datasets. A home study of 32 T1D adolescents in a free-living condition was conducted. Simultaneously recorded accelerometer and continuous glucose monitor (CGM) data of 194.6 h were collected from 15 participants. These data were utilized for training and testing the predictive model. Four lengths of the sliding window (1, 2, 5, and 10 s) and four machine learning algorithms (decision tree, support vector machine, k-nearest neighbor, and bagged trees ensemble classifier) were applied to classify tremor as non-hypoglycemic or hypoglycemic. The greatest accuracy (86.65%) was achieved by the ensemble classifier (bagged trees based on the subspace k-NN) for 1 s window length. Multiple prediction windows are merged to generate aggregate sequential predictions. Prediction accuracy of 86.05% was achieved for a 15 s batch (3 s window length). This study demonstrates the feasibility of detecting and predicting the onset of hypoglycemia based on hand tremor data collected by wrist-worn accelerometer sensors. [ABSTRACT FROM AUTHOR]

Most incidents in complex systems such as process plants are not-chance events and weak signals emerge for a long time before incidents occur. It is necessary to identify and respond to the weak signals as early as possible for preventing incidents. However, in the era of Industry 4.0, recognizing weak signals from the abundance of data is challenging. Since the terminology "weak signal" was not precisely defined, the study first proposed a formal definition of "weak signal" and discussed its characteristics. Additionally, a framework was developed to address the challenges of observing, evaluating, and responding to weak signals in complex systems. The framework first utilized Function Resonance Analysis Method (FRAM) to determine the information to be collected for observing weak signals. Then, the relevance of weak signals and the corresponding responses were evaluated by utilizing Balanced Random Forest (BRF), Weighted Random Forest (WRF) and Decision Tree (DT) classification models. The case study of a hypothetical batch process showed great potentials for applying the framework in real operations. Based on the potential weak signals that were indicated by FRAM, probabilities of temperature deviations in the process were predicted with high accuracy by the optimal BRF model. Underlying weak signals and the corresponding responses to reduce the probabilities were identified from the DT. [ABSTRACT FROM AUTHOR]

Most incidents in complex systems such as process plants are not-chance events and weak signals emerge for a long time before incidents occur. It is necessary to identify and respond to the weak signals as early as possible for preventing incidents. However, in the era of Industry 4.0, recognizing weak signals from the abundance of data is challenging. Since the terminology "weak signal" was not precisely defined, the study first proposed a formal definition of "weak signal" and discussed its characteristics. Additionally, a framework was developed to address the challenges of observing, evaluating, and responding to weak signals in complex systems. The framework first utilized Function Resonance Analysis Method (FRAM) to determine the information to be collected for observing weak signals. Then, the relevance of weak signals and the corresponding responses were evaluated by utilizing Balanced Random Forest (BRF), Weighted Random Forest (WRF) and Decision Tree (DT) classification models. The case study of a hypothetical batch process showed great potentials for applying the framework in real operations. Based on the potential weak signals that were indicated by FRAM, probabilities of temperature deviations in the process were predicted with high accuracy by the optimal BRF model. Underlying weak signals and the corresponding responses to reduce the probabilities were identified from the DT. [ABSTRACT FROM AUTHOR]