An autonomous channel deep learning framework for blood glucose prediction

After-meal blood glucose level prediction for type-2 diabetic patients

Type 2 Diabetes, a metabolic disorder disease, is becoming a fast growing health crisis worldwide. It reduces the quality of life, and increases mortality and health care costs unless managed well. After-meal blood glucose level measure is considered as one of the most fundamental and well-recognized steps in managing Type 2 diabetes as it guides a user to make better plans of their diet and thus control the diabetes well. In this paper, we propose a data-driven approach to predict the 2 h after meal blood glucose level from the previous discrete blood glucose readings, meal, exercise, medication, & profile information of Type 2 diabetes patients. To the best of our knowledge, this is the first attempt to use discrete blood glucose readings for 2 h after meal blood glucose level prediction using data-driven models. In this study, we have collected data from five prediabetic and diabetic patients in free living conditions for six months. We have presented comparative experimental study using different popular machine learning models including support vector regression, random forest, and extreme gradient boosting, and two deep layer techniques: multilayer perceptron, and convolutional neural network. We present also the impact of different features in blood glucose level prediction, where we observe that meal has some modest and medication has a good influence on blood glucose level.

AWD-stacking: An enhanced ensemble learning model for predicting glucose levels

Accurate prediction of blood glucose levels is essential for type 1 diabetes optimizing insulin therapy and minimizing complications in patients with type 1 diabetes. Using ensemble learning algorithms is a promising approach. In this regard, this study proposes an improved stacking ensemble learning algorithm for predicting blood glucose level, in which three improved long short-term memory network models are used as the base model, and an improved nearest neighbor propagation clustering algorithm is adaptively weighted to this ensemble model. The OhioT1DM dataset is used to train and evaluate the performance of the proposed model. This study evaluated the performance of the proposed model using the Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Matthews Correlation Coefficient (MCC) as the evaluation metrics. The experimental results demonstrate that the proposed model achieves an RMSE of 1.425 mg/dL, MAE of 0.721 mg/dL, and MCC of 0.982 mg/dL for a 30-minute prediction horizon(PH), RMSE of 3.212 mg/dL, MAE of 1.605 mg/dL, and MCC of 0.950 mg/dL for a 45-minute PH; and RMSE of 6.346 mg/dL, MAE of 3.232 mg/dL, and MCC of 0.930 mg/dL for a 60-minute PH. Compared with the best non-ensemble model StackLSTM, the RMSE and MAE were improved by up to 27.92% and 65.32%, respectively. Clarke Error Grid Analysis and critical difference diagram revealed that the model errors were within 10%. The model proposed in this study exhibits state-of-the-art predictive performance, making it suitable for clinical decision-making and of significant importance for the effective treatment of diabetes in patients.

A personalized multitasking framework for real-time prediction of blood glucose levels in type 1 diabetes patients

Real-time prediction of blood glucose levels (BGLs) in individuals with type 1 diabetes (T1D) presents considerable challenges. Accordingly, we present a personalized multitasking framework aimed to forecast blood glucose levels in patients. The patient data was initially categorized according to gender and age and subsequently utilized as input for a modified GRU network model, creating five prediction sub-models. The model hyperparameters were optimized and tuned after introducing the decay factor and incorporating the TCN network and attention mechanism into the GRU model. This step was undertaken to improve the capability of feature extraction. The Ohio T1DM clinical dataset was used to train and evaluate the performance of the proposed model. The metrics, including Root Mean Square Error (RMSE), Mean Absolute Error (MAE) and Clark Error Grid Analysis (EGA), were used to evaluate the performance. The results showed that the average RMSE and the MAE of the proposed model were 16.896 and 9.978 mg/dL, respectively, over the prediction horizon (PH) of 30 minutes. The average RMSE and the MAE were 28.881 and 19.347 mg/dL, respectively, over the PH of 60 min. The proposed model demonstrated excellent prediction accuracy. In addition, the EGA analysis showed that the proposed model accurately predicted 30-minute and 60-minute PH within zones A and B, demonstrating that the framework is clinically feasible. The proposed personalized multitask prediction model in this study offers robust assistance for clinical decision-making, playing a pivotal role in improving the outcomes of individuals with diabetes.

A 2-dimensional model framework for blood glucose prediction based on iterative learning control architecture

The accurate, timely, and personalized prediction for future blood glucose (BG) levels is undoubtedly needed for further advancement of diabetes management technologies. Human inherent circadian rhythm and regular lifestyle resulting in similarity of daily glycemic dynamics play a positive role in the prediction of blood glucose. Inspired by the iterative learning control (ILC) method in the field of automatic control, a 2-dimensional (2-D) model framework is constructed to predict the future blood glucose levels by taking both the short-range information within a day (intra-day) and long-range information between days (inter-day) into account. In this framework, the radial basis function neural network was applied to capture nonlinear relationships in glycemic metabolism, that is, short-range temporal dependence and long-range contemporaneous dependence on previous days. We build models for each patient, and the models were tested on the in silico datasets at various prediction horizons (PHs). The learning model developed in the 2-D framework successfully increases the accuracy and reduces the delay of predictions. This modeling framework provides a new point of view for BG level prediction and contributes to the development of personalized glucose management, such as hypoglycemia warning and glycemic control.

Unsupervised dual-domain disentangled network for removal of rigid motion artifacts in MRI

Motion artifacts in magnetic resonance imaging (MRI) have always been a serious issue because they can affect subsequent diagnosis and treatment. Supervised deep learning methods have been investigated for the removal of motion artifacts; however, they require paired data that are difficult to obtain in clinical settings. Although unsupervised methods are widely proposed to fully use clinical unpaired data, they generally focus on anatomical structures generated by the spatial domain while ignoring phase error (deviations or inaccuracies in phase information that are possibly caused by rigid motion artifacts during image acquisition) provided by the frequency domain. In this study, a 2D unsupervised deep learning method named unsupervised disentangled dual-domain network (UDDN) was proposed to effectively disentangle and remove unwanted rigid motion artifacts from images. In UDDN, a dual-domain encoding module was presented to capture different types of information from the spatial and frequency domains to enrich the information. Moreover, a cross-domain attention fusion module was proposed to effectively fuse information from different domains, reduce information redundancy, and improve the performance of motion artifact removal. UDDN was validated on a publicly available dataset and a clinical dataset. Qualitative and quantitative experimental results showed that our method could effectively remove motion artifacts and reconstruct image details. Moreover, the performance of UDDN surpasses that of several state-of-the-art unsupervised methods and is comparable with that of the supervised method. Therefore, our method has great potential for clinical application in MRI, such as real-time removal of rigid motion artifacts.

Heterogeneous temporal representation for diabetic blood glucose prediction

Background and aims: Blood glucose prediction (BGP) has increasingly been adopted for personalized monitoring of blood glucose levels in diabetic patients, providing valuable support for physicians in diagnosis and treatment planning. Despite the remarkable success achieved, applying BGP in multi-patient scenarios remains problematic, largely due to the inherent heterogeneity and uncertain nature of continuous glucose monitoring (CGM) data obtained from diverse patient profiles. Methodology: This study proposes the first graph-based Heterogeneous Temporal Representation (HETER) network for multi-patient Blood Glucose Prediction (BGP). Specifically, HETER employs a flexible subsequence repetition method (SSR) to align the heterogeneous input samples, in contrast to the traditional padding or truncation methods. Then, the relationships between multiple samples are constructed as a graph and learned by HETER to capture global temporal characteristics. Moreover, to address the limitations of conventional graph neural networks in capturing local temporal dependencies and providing linear representations, HETER incorporates both a temporally-enhanced mechanism and a linear residual fusion into its architecture. Results: Comprehensive experiments were conducted to validate the proposed method using real-world data from 112 patients in two hospitals, comparing it with five well-known baseline methods. The experimental results verify the robustness and accuracy of the proposed HETER, which achieves the maximal improvement of 31.42%, 27.18%, and 34.85% in terms of MAE, MAPE, and RMSE, respectively, over the second-best comparable method. Discussions: HETER integrates global and local temporal information from multi-patient samples to alleviate the impact of heterogeneity and uncertainty. This method can also be extended to other clinical tasks, thereby facilitating efficient and accurate capture of crucial pattern information in structured medical data.

Jointly Predicting Postprandial Hypoglycemia and Hyperglycemia Using Continuous Glucose Monitoring Data in Type 1 Diabetes

The development of continuous glucose monitoring (CGM) systems has enabled people with type 1 diabetes mellitus (T1DM) to track their glucose trajectory in real-time and inspired research in personalised glucose prediction. In this paper, our aim is to predict postprandial abnormal-glycemia events. Different from prior research which focuses on hypoglycemia only, we make the first attempt to establish our problem as the joint prediction of hyperglycemia and hypoglycemia. On this basis, we propose a machine learning model that learns from the pattern of 1 hour past glucose and makes predictions for the two tasks simultaneously using a unified backbone. Key benefits of our methodology include 1) requiring only the CGM sequence as the input, thus making it more widely applicable than other counterparts using extra inputs such as the nutrition details, and 2) minimising the computational cost as the two tasks are unified into a single model. Our experiments on the openly available OhioT1DM dataset achieve state-of-the-art performance (Matthew's correlation coefficient of 0.61 for hyperglycemia and 0.48 for hypoglycemia). To encourage further study, we release our codes at https://github.com/r-cui/PostprandialHyperHypoPrediction under the MIT license.

Blood Glucose Level Time Series Forecasting: Nested Deep Ensemble Learning Lag Fusion

Blood glucose level prediction is a critical aspect of diabetes management. It enables individuals to make informed decisions about their insulin dosing, diet, and physical activity. This, in turn, improves their quality of life and reduces the risk of chronic and acute complications. One conundrum in developing time-series forecasting models for blood glucose level prediction is to determine an appropriate length for look-back windows. On the one hand, studying short histories foists the risk of information incompletion. On the other hand, analysing long histories might induce information redundancy due to the data shift phenomenon. Additionally, optimal lag lengths are inconsistent across individuals because of the domain shift occurrence. Therefore, in bespoke analysis, either optimal lag values should be found for each individual separately or a globally suboptimal lag value should be used for all. The former approach degenerates the analysis's congruency and imposes extra perplexity. With the latter, the fine-tunned lag is not necessarily the optimum option for all individuals. To cope with this challenge, this work suggests an interconnected lag fusion framework based on nested meta-learning analysis that improves the accuracy and precision of predictions for personalised blood glucose level forecasting. The proposed framework is leveraged to generate blood glucose prediction models for patients with type 1 diabetes by scrutinising two well-established publicly available Ohio type 1 diabetes datasets. The models developed undergo vigorous evaluation and statistical analysis from mathematical and clinical perspectives. The results achieved underpin the efficacy of the proposed method in blood glucose level time-series prediction analysis.

Long-Term Glucose Forecasting for Open-Source Automated Insulin Delivery Systems: A Machine Learning Study with Real-World Variability Analysis

Glucose forecasting serves as a backbone for several healthcare applications, including real-time insulin dosing in people with diabetes and physical activity optimization. This paper presents a study on the use of machine learning (ML) and deep learning (DL) methods for predicting glucose variability (GV) in individuals with open-source automated insulin delivery systems (AID). A three-stage experimental framework is employed in this work to systematically implement and evaluate ML/DL methods on a large-scale diabetes dataset collected from individuals with open-source AID. The first stage involves data collection, the second stage involves data preparation and exploratory analysis, and the third stage involves developing, fine-tuning, and evaluating ML/DL models. The performance and resource costs of the models are evaluated alongside relative and proportional errors for 17 GV metrics. Evaluation of fine-tuned ML/DL models shows considerable accuracy in glucose forecasting and variability analysis up to 48 h in advance. The average MAE ranges from 2.50 mg/dL for long short-term memory models (LSTM) to 4.94 mg/dL for autoregressive integrated moving average (ARIMA) models, and the RMSE ranges from 3.7 mg/dL for LSTM to 7.67 mg/dL for ARIMA. Model execution time is proportional to the amount of data used for training, with long short-term memory models having the lowest execution time but the highest memory consumption compared to other models. This work successfully incorporates the use of appropriate programming frameworks, concurrency-enhancing tools, and resource and storage cost estimators to encourage the sustainable use of ML/DL in real-world AID systems.

Prediction of Blood Risk Score in Diabetes Using Deep Neural Networks

Improving the prediction of blood glucose concentration may improve the quality of life of people living with type 1 diabetes by enabling them to better manage their care. Given the anticipated benefits of such a prediction, numerous methods have been proposed. Rather than attempting to predict glucose concentration, a deep learning framework for prediction is proposed in which prediction is performed using a scale for hypo- and hyper-glycemia risk. Using the blood glucose risk score formula proposed by Kovatchev et al., models with different architectures were trained, including, a recurrent neural network (RNN), a gated recurrent unit (GRU), a long short-term memory (LSTM) network, and an encoder-like convolutional neural network (CNN). The models were trained using the OpenAPS Data Commons data set, comprising 139 individuals, each with tens of thousands of continuous glucose monitor (CGM) data points. The training set was composed of 7% of the data set, while the remaining was used for testing. Performance comparisons between the different architectures are presented and discussed. To evaluate these predictions, performance results are compared with the last measurement (LM) prediction, through a sample-and-hold approach continuing the last known measurement forward. The results obtained are competitive when compared to other deep learning methods. A root mean squared error (RMSE) of 16 mg/dL, 24 mg/dL, and 37 mg/dL were obtained for CNN prediction horizons of 15, 30, and 60 min, respectively. However, no significant improvements were found for the deep learning models compared to LM prediction. Performance was found to be highly dependent on architecture and the prediction horizon. Lastly, a metric to assess model performance by weighing each prediction point error with the corresponding blood glucose risk score is proposed. Two main conclusions are drawn. Firstly, going forward, there is a need to benchmark model performance using LM prediction to enable the comparison between results obtained from different data sets. Secondly, model-agnostic data-driven deep learning models may only be meaningful when combined with mechanistic physiological models; here, it is argued that neural ordinary differential equations may combine the best of both approaches. These findings are based on the OpenAPS Data Commons data set and are to be validated in other independent data sets.