A deep learning approach to adherence detection for type 2 diabetics

Performance analysis and prediction of type 2 diabetes mellitus based on lifestyle data using machine learning approaches

Objective

Diabetes is a chronic fatal disease that has affected millions of people all over the globe. Type 2 Diabetes Mellitus (T2DM) accounts for 90% of the affected population among all types of diabetes. Millions of T2DM patients remain undiagnosed due to lack of awareness and under resourced healthcare system. So, there is a dire need for a diagnostic and prognostic tool that shall help the healthcare providers, clinicians and practitioners with early prediction and hence can recommend the lifestyle changes required to stop the progression of diabetes. The main objective of this research is to develop a framework based on machine learning techniques using only lifestyle indicators for prediction of T2DM disease. Moreover, prediction model can be used without visiting clinical labs and hospital readmissions.

Method

A proposed framework is presented and implemented based on machine learning paradigms using lifestyle indicators for better prediction of T2DM disease. The current research has involved different experts like Diabetologists, Endocrinologists, Dieticians, Nutritionists, etc. for selecting the contributing 1552 instances and 11 attributes lifestyle biological features to promote health and manage complications towards T2DM disease. The dataset has been collected through survey and google forms from different geographical regions.

Results

Seven machine learning classifiers were employed namely K-Nearest Neighbour (KNN), Linear Regression (LR), Support Vector Machine (SVM), Naive Bayes (NB), Decision Tree (DT), Random Forest (RF) and Gradient Boosting (GB). Gradient Boosting classifier outperformed best with an accuracy rate of 97.24% for training and 96.90% for testing separately followed by RF, DT, NB, SVM, LR, and KNN as 95.36%, 92.52%, 90.72%, 90.20%, 90.20% and 77.06% respectively. However, in terms of precision, RF achieved high performance (0.980%) and KNN performed the lowest (0.793%). As far as recall is being concerned, GB achieved the highest rate of 0.975% and KNN showed the worst rate of 0.774%. Also, GB is top performed in terms of f1-score. According to the ROCs, GB and NB had a better area under the curve compared to the others.

Conclusion

The research developed a realistic health management system for T2DM disease based on machine learning techniques using only lifestyle data for prediction of T2DM. To extend the current study, these models shall be used for different, large and real-time datasets which share the commonality of data with T2DM disease to establish the efficacy of the proposed system.

A Comprehensive Review of Various Diabetic Prediction Models: A Literature Survey

Diabetes is a chronic disease characterized by a high amount of glucose in the blood and can cause too many complications also in the body, such as internal organ failure, retinopathy, and neuropathy. According to the predictions made by WHO, the figure may reach approximately 642 million by 2040, which means one in a ten may suffer from diabetes due to unhealthy lifestyle and lack of exercise. Many authors in the past have researched extensively on diabetes prediction through machine learning algorithms. The idea that had motivated us to present a review of various diabetic prediction models is to address the diabetic prediction problem by identifying, critically evaluating, and integrating the findings of all relevant, high-quality individual studies. In this paper, we have analysed the work done by various authors for diabetes prediction methods. Our analysis on diabetic prediction models was to find out the methods so as to select the best quality researches and to synthesize the different researches. Analysis of diabetes data disease is quite challenging because most of the data in the medical field are nonlinear, nonnormal, correlation structured, and complex in nature. Machine learning-based algorithms have been ruled out in the field of healthcare and medical imaging. Diabetes mellitus prediction at an early stage requires a different approach from other approaches. Machine learning-based system risk stratification can be used to categorize the patients into diabetic and controls. We strongly recommend our study because it comprises articles from various sources that will help other researchers on various diabetic prediction models.

A Smart Healthcare Recommendation System for Multidisciplinary Diabetes Patients with Data Fusion Based on Deep Ensemble Learning

The prediction of human diseases precisely is still an uphill battle task for better and timely treatment. A multidisciplinary diabetic disease is a life-threatening disease all over the world. It attacks different vital parts of the human body, like Neuropathy, Retinopathy, Nephropathy, and ultimately Heart. A smart healthcare recommendation system predicts and recommends the diabetic disease accurately using optimal machine learning models with the data fusion technique on healthcare datasets. Various machine learning models and methods have been proposed in the recent past to predict diabetes disease. Still, these systems cannot handle the massive number of multifeatures datasets on diabetes disease properly. A smart healthcare recommendation system is proposed for diabetes disease based on deep machine learning and data fusion perspectives. Using data fusion, we can eliminate the irrelevant burden of system computational capabilities and increase the proposed system's performance to predict and recommend this life-threatening disease more accurately. Finally, the ensemble machine learning model is trained for diabetes prediction. This intelligent recommendation system is evaluated based on a well-known diabetes dataset, and its performance is compared with the most recent developments from the literature. The proposed system achieved 99.6% accuracy, which is higher compared to the existing deep machine learning methods. Therefore, our proposed system is better for multidisciplinary diabetes disease prediction and recommendation. Our proposed system's improved disease diagnosis performance advocates for its employment in the automated diagnostic and recommendation systems for diabetic patients.

Predicting Progression Patterns of Type 2 Diabetes using Multi-sensor Measurements

Type 2 diabetes - a prevalent chronic disease worldwide - increases risk for serious health consequences including heart and kidney disease. Forecasting diabetes progression can inform disease management strategies, thereby potentially reducing the likelihood or severity of its consequences. We use continuous glucose monitoring and actigraphy data from 54 individuals with Type 2 diabetes to predict their future hemoglobin A1c, HDL cholesterol, LDL cholesterol, and triglyceride levels one year later. We use a combination of convolutional and recurrent neural networks to develop a deep neural network architecture that can learn the dynamic patterns in different sensors' data and combine those patterns with additional demographic and lab data. To further demonstrate the generalizability of our models, we also evaluate their performance using an independent public dataset of individuals with Type 1 diabetes. In addition to diabetes, our approach could be useful for other serious and chronic physical illness, where dynamic (e.g., from multiple sensors) and static (e.g., demographic) data are used for creating predictive models.

Predicting medication adherence using ensemble learning and deep learning models with large scale healthcare data

Clinical studies from WHO have demonstrated that only 50–70% of patients adhere properly to prescribed drug therapy. Such adherence failure can impact therapeutic efficacy for the patients in question and compromises data quality around the population-level efficacy of the drug for the indications targeted. In this study, we applied various ensemble learning and deep learning models to predict medication adherence among patients. Our contribution to this endeavour involves targeting the problem of adherence prediction for a particularly challenging class of patients who self-administer injectable medication at home. Our prediction pipeline, based on event history, comprises a connected sharps bin which aims to help patients better manage their condition and improve outcomes. In other words, the efficiency of interventions can be significantly improved by prioritizing the patients who are most likely to be non-adherent. The collected data comprising a rich event feature set may be exploited for the purposes of predicting the status of the next adherence state for individual patients. This paper reports on how this concept can be realized through an investigation using a wide range of ensemble learning and deep learning models on a real-world dataset collected from such a system. The dataset investigated comprises 342,174 historic injection disposal records collected over the course of more than 5 years. A comprehensive comparison of different models is given in this paper. Moreover, we demonstrate that the selected best performer, long short-term memory (LSTM), generalizes well by deploying it in a true future testing dataset. The proposed end-to-end pipeline is capable of predicting patient failure in adhering to their therapeutic regimen with 77.35 % accuracy (Specificity: 78.28 %, Sensitivity: 76.42%, Precision: 77.87%,F1 score: 0.7714, ROC AUC: 0.8390).

Machine Learning-Based Adherence Detection of Type 2 Diabetes Patients on Once-Daily Basal Insulin Injections

BACKGROUND

Lack of treatment adherence can lead to life-threatening health complications for people with type 2 diabetes (T2D). Recent improvements and availability in continuous glucose monitoring (CGM) technology have enabled various possibilities to monitor diabetes treatment. Detection of missed once-daily basal insulin injections can be used to provide feedback to patients, thus improving their diabetes management. In this study, we explore how machine learning (ML) based on CGM data can be used for detecting adherence to once-daily basal insulin injections.

METHODS

In-silico CGM data were generated to simulate a cohort of T2D patients on once-daily insulin injection (Tresiba®). Deep learning methods within ML based on automatic feature extraction including convolutional neural networks were explored and compared with simple feature-engineered ML classification models for adherence detection. It was further investigated whether fused expert-dependent and automatically learned features could improve performance, resulting in a comparison of six different detection models. Adherence was detected throughout each day with an increasing amount of CGM data available.

RESULTS

The adherence detection accuracy improved as more CGM data became available on the day of classification. The three classification models based on expert-engineered features obtained mean accuracies of 78.6%, 78.2%, and 78.3%. The classification model based purely on learned features obtained a mean accuracy of 79.7%. The two classification models fusing expert-engineered and learned features obtained mean accuracies of 79.7% and 79.8%. All the mentioned results were obtained 16 hours after time of injection.

CONCLUSION

The results suggest that adherence detection based on CGM data is feasible. Even though our study based on in-silico data indicates only slightly improved performance of more complex models, the question remains whether advanced models would outperform the simple in a real-world setting. Thus, future studies on adherence monitoring using real CGM data are relevant.

Use of Machine Learning Approaches in Clinical Epidemiological Research of Diabetes

PURPOSE OF REVIEW

Machine learning approaches-which seek to predict outcomes or classify patient features by recognizing patterns in large datasets-are increasingly applied to clinical epidemiology research on diabetes. Given its novelty and emergence in fields outside of biomedical research, machine learning terminology, techniques, and research findings may be unfamiliar to diabetes researchers. Our aim was to present the use of machine learning approaches in an approachable way, drawing from clinical epidemiological research in diabetes published from 1 Jan 2017 to 1 June 2020.

RECENT FINDINGS

Machine learning approaches using tree-based learners-which produce decision trees to help guide clinical interventions-frequently have higher sensitivity and specificity than traditional regression models for risk prediction. Machine learning approaches using neural networking and "deep learning" can be applied to medical image data, particularly for the identification and staging of diabetic retinopathy and skin ulcers. Among the machine learning approaches reviewed, researchers identified new strategies to develop standard datasets for rigorous comparisons across older and newer approaches, methods to illustrate how a machine learner was treating underlying data, and approaches to improve the transparency of the machine learning process. Machine learning approaches have the potential to improve risk stratification and outcome prediction for clinical epidemiology applications. Achieving this potential would be facilitated by use of universal open-source datasets for fair comparisons. More work remains in the application of strategies to communicate how the machine learners are generating their predictions.

Short Term Blood Glucose Prediction based on Continuous Glucose Monitoring Data

Continuous Glucose Monitoring (CGM) has enabled important opportunities for diabetes management. This study explores the use of CGM data as input for digital decision support tools. We investigate how Recurrent Neural Networks (RNNs) can be used for Short Term Blood Glucose (STBG) prediction and compare the RNNs to conventional time-series forecasting using Autoregressive Integrated Moving Average (ARIMA). A prediction horizon up to 90 min into the future is considered. In this context, we evaluate both population-based and patient-specific RNNs and contrast them to patient-specific ARIMA models and a simple baseline predicting future observations as the last observed. We find that the population-based RNN model is the best performing model across the considered prediction horizons without the need of patient-specific data. This demonstrates the potential of RNNs for STBG prediction in diabetes patients towards detecting/mitigating severe events in the STBG, in particular hypoglycemic events. However, further studies are needed in regards to the robustness and practical use of the investigated STBG prediction models.

Fast Assessment of Glycemic Control based on Continuous Glucose Monitoring Data

Diabetes has become a major public health problem in the world. In this context, early assessment of glycemic control is essential in order to avoid life-threatening health complications. A panel of diabetes experts have recently proposed a list of recommendations when using Continuous Glucose Monitoring (CGM) for glycemic control assessment including a minimum of two weeks of CGM data. A recent study has further introduced a metric called Glucose Profile Indicator (GPI) for CGM based diabetes management including a subset of the recommended CGM metrics. In this pilot study, it was investigated if less than two weeks of CGM data would impact the performance of GPI compared to the proposed two weeks of CGM data. Furthermore, logistic regression (LR) was used to examine if an improvement could be achieved taking as input the CGM metrics used to quantify GPI. The population mean accuracy for accumulated day 1 to 13 varied between 72.8 ± 2.0% - 98.3 ± 0.4% with no clear sign of improvement using LR. Hence, this indicates a trade-off between the amount of available CGM data and the precision in which the GPI outcome using all 14 days can be achieved when considering features of the GPI alone. Future work is needed to investigate if this trade-off can be improved by the use of additional features of the CGM.

Deep Learning on 1-D Biosignals: a Taxonomy-based Survey

OBJECTIVES

 Deep learning models such as convolutional neural networks (CNNs) have been applied successfully to medical imaging, but biomedical signal analysis has yet to fully benefit from this novel approach. Our survey aims at (i) reviewing deep learning techniques for biosignal analysis in computer- aided diagnosis; and (ii) deriving a taxonomy for organizing the growing number of applications in the field.

METHODS

 A comprehensive literature research was performed using PubMed, Scopus, and ACM. Deep learning models were classified with respect to the (i) origin, (ii) dimension, and (iii) type of the biosignal as input to the deep learning model; (iv) the goal of the application; (v) the size and (vi) type of ground truth data; (vii) the type and (viii) schedule of learning the network; and (ix) the topology of the model.

RESULTS

 Between January 2010 and December 2017, a total 71 papers were published on the topic. The majority (n = 36) of papers are on electrocariography (ECG) signals. Most applications (n = 25) aim at detection of patterns, while only a few (n = 6) at predection of events. Out of 36 ECG-based works, many (n = 17) relate to multi-lead ECG. Other biosignals that have been identified in the survey are electromyography, phonocardiography, photoplethysmography, electrooculography, continuous glucose monitoring, acoustic respiratory signal, blood pressure, and electrodermal activity signal, while ballistocardiography or seismocardiography have yet to be analyzed using deep learning techniques. In supervised and unsupervised applications, CNNs and restricted Boltzmann machines are the most and least frequently used, (n = 34) and (n = 15), respectively.

CONCLUSION

 Our key-code classification of relevant papers was used to cluster the approaches that have been published to date and demonstrated a large variability of research with respect to data, application, and network topology. Future research is expected to focus on the standardization of deep learning architectures and on the optimization of the network parameters to increase performance and robustness. Furthermore, application-driven approaches and updated training data from mobile recordings are needed.