Diabetes Detection and Management through Photoplethysmographic and Electrocardiographic Signals Analysis: A Systematic Review

(1) Background: Diabetes mellitus (DM) is a chronic, metabolic disease characterized by elevated levels of blood glucose. Recently, some studies approached the diabetes care domain through the analysis of the modifications of cardiovascular system parameters. In fact, cardiovascular diseases are the first leading cause of death in diabetic subjects. Thanks to their cost effectiveness and their ease of use, electrocardiographic (ECG) and photoplethysmographic (PPG) signals have recently been used in diabetes detection, blood glucose estimation and diabetes-related complication detection. This review's aim is to provide a detailed overview of all the published methods, from the traditional (non machine learning) to the deep learning approaches, to detect and manage diabetes using PPG and ECG signals. This review will allow researchers to compare and understand the differences, in terms of results, amount of data and complexity that each type of approach provides and requires. (2) Method: We performed a systematic review based on articles that focus on the use of ECG and PPG signals in diabetes care. The search was focused on keywords related to the topic, such as "Diabetes", "ECG", "PPG", "Machine Learning", etc. This was performed using databases, such as PubMed, Google Scholar, Semantic Scholar and IEEE Xplore. This review's aim is to provide a detailed overview of all the published methods, from the traditional (non machine learning) to the deep learning approaches, to detect and manage diabetes using PPG and ECG signals. This review will allow researchers to compare and understand the differences, in terms of results, amount of data and complexity that each type of approach provides and requires. (3) Results: A total of 78 studies were included. The majority of the selected studies focused on blood glucose estimation (41) and diabetes detection (31). Only 7 studies focused on diabetes complications detection. We present these studies by approach: traditional, machine learning and deep learning approaches. (4) Conclusions: ECG and PPG analysis in diabetes care showed to be very promising. Clinical validation and data processing standardization need to be improved in order to employ these techniques in a clinical environment.

Analysis for warning factors of type 2 diabetes mellitus complications with Markov blanket based on a Bayesian network model

BACKGROUND AND OBJECTIVE

Type 2 diabetes mellitus (T2DM) complications seriously affect the quality of life and could not be cured completely. Actions should be taken for prevention and self-management. Analysis of warning factors is beneficial for patients, on which some previous studies focused. They generally used the professional medical test factors or complete factors to predict and prevent, but it was inconvenient and impractical for patients to self-manage. With this in mind, this study built a Bayesian network (BN) model, from the perspective of diabetic patients' self-management and prevention, to predict six complications of T2DM using the selected warning factors which patients could have access from medical examination. Furthermore, the model was analyzed to explore the relationships between physiological variables and T2DM complications, as well as the complications themselves. The model aims to help patients with T2DM self-manage and prevent themselves from complications.

METHODS

The dataset was collected from a well-known data center called the National Health Clinical Center between 1st January 2009 and 31st December 2009. After preprocess and impute the data, a BN model merging expert knowledge was built with Bootstrap and Tabu search algorithm. Markov Blanket (MB) was used to select the warning factors and predict T2DM complications. Moreover, a Bayesian network without prior information (BN-wopi) model learned using 10-fold cross-validation both in structure and in parameters was added to compare with other classifiers learned using 10-fold cross-validation fairly. The warning factors were selected according the structure learned in each fold and were used to predict. Finally, the performance of two BN models using warning features were compared with Naïve Bayes model, Random Forest model, and C5.0 Decision Tree model, which used all features to predict. Besides, the validation parameters of the proposed model were also compared with those in existing studies using some other variables in clinical data or biomedical data to predict T2DM complications.

RESULTS

Experimental results indicated that the BN models using warning factors performed statistically better than their counterparts using all other variables in predicting T2DM complications. In addition, the proposed BN model were effective and significant in predicting diabetic nephropathy (DN) (AUC: 0.831), diabetic foot (DF) (AUC: 0.905), diabetic macrovascular complications (DMV) (AUC: 0.753) and diabetic ketoacidosis (DK) (AUC: 0.877) with the selected warning factors compared with other experiments.

CONCLUSIONS

The warning factors of DN, DF, DMV, and DK selected by MB in this research might be able to help predict certain T2DM complications effectively, and the proposed BN model might be used as a general tool for prevention, monitoring, and self-management.

Extended Central Tendency Measure and difference plot for heart rate variability analysis

Heart rate variability (HRV) is a non-invasive alternative to analyze the role of the autonomic nervous system (ANS) on heart functioning. Many tools have been developed to analyze collected cardiac data. Among them, the Central Tendency Measure (CTM) is a quantitative method for variability analysis of RR intervals. The values of the CTM must be between 0 and 1 (inclusive) for different radius, which follows the intrinsic characteristics of each time series. Using the conventional CTM, the successive differences of the time series may be calculated, and it can classify and differentiate the disturbances in the ANS involving HRV. This method was extended (e-CTM) to analyze the differences between RR interval time series. In this extension, a new parameter is added, which allows analysis of long time intervals, instead of successive and adjacent RR intervals. The ability of the e-CTM to differentiate the groups of the RR interval time series was verified with 145 RR interval time series divided into three groups: subjects with congestive heart failure, healthy subjects, and nurses during one hour of their workday. Results evidence that the new parameter added differentiates the group with pathology (and subsequent impairment of ANS) and group under stress at work (temporary impairment of ANS). These results suggest that the e-CTM is capable of detection long-term variations in the HRV according to the ANS impairment.

ASSESSMENT OF CARDIAC HEART FAILURE AND CARDIAC ARTERY DISEASE BY THE HIGHER ORDER SPECTRA

Cardiac diseases are major reason of death in the world populace and the numeral of cases is upsurging every year. Due to cardiac artery disease (CAD), the strength of heart muscles becomes weak and heart pumping is disturbed which may eventually lead to abnormal heart beat and heart failure. Therefore, the beginning stage detection of CAD and cardiac heart failure (CHF) are of prime importance. In this work, we have used a non-invasive diagnosis method as higher order spectra (HOS) for assessment of cardiac diseases. The method indicates whether or not a cardiac heart disease is present, by assessing the cardiac health of subjects using extracted features from heart rate variability (HRV) signals. This assessment is based on 10 spectra nonlinear features. These features were extracted from HRV signals by using the HOS method. For this study, the R-R interval data (i.e. HRV signals) were taken from the standard database of cardiac heart failure (CHF), CAD patients, healthy young (YNG) and Self recorded of healthy young (SELF_YNG) subjects. Statistical assessments were performed on the group of database sets as YNG-CAD, YNG-CHF, SELF_YNG-CAD and Self_YNG-CHF subjects. A Wilcoxon rank sum test ([Formula: see text]-value) was used to statistically compare the features extracted by HOS for group of data sets. It indicates whether or not the same features of individual classes of HRV data sets are dissimilar. The results depicted that the all features are very significant ([Formula: see text]) except the phase entropy (PHE) feature which is not significant for CAD-CHF, SELF_YNG-CAD and SELF_YNG-CHF group of subjects. While in the case of YNG-CAD group of subjects, features like first-order spectral moment of amplitudes of diagonal elements (H3), PHE and logarithmic amplitudes of diagonal elements (H2) are significant ([Formula: see text]) and excluding these features, the remaining features are very significant except MM and H1 which are not significant. The results also depicted that the mean value of sum of logarithmic amplitude (H1), H2, normalized entropy (P1), normalized squared entropy (P2) and PHE features of healthy YNG subjects are having higher values than that of CAD and CHF patients. While weighted center of bi-spectrum (WCOB2) and FLAT spectrum features are lower than CAD and CHF patients compared to YNG subjects. In case of CAD and CHF patients, all the features of CAD patients are having higher values compared to CHF except P1, P2 and WCOB1.