A principal component regression approach for estimation of ventricular repolarization characteristics

Non-invasive method for blood glucose monitoring using ECG signal

Introduction: Tight glucose monitoring is crucial for diabetic patients by using a Continuous Glucose Monitor (CGM). The existing CGMs measure the Blood Glucose Concentration (BGC) from the interstitial fluid. These technologies are quite expensive, and most of them are invasive. Previous studies have demonstrated that hypoglycemia and hyperglycemia episodes affect the electrophysiology of the heart. However, they did not determine a cohort relationship between BGC and ECG parameters.

Material and method: In this work, we propose a new method for determining the BGC using surface ECG signals. Recurrent Convolutional Neural Networks (RCNN) were applied to segment the ECG signals. Then, the extracted features were employed to determine the BGC using two mathematical equations. This method has been tested on 04 patients over multiple days from the D1namo dataset, using surface ECG signals instead of intracardiac signal.

Results: We were able to segment the ECG signals with an accuracy of 94% using the RCNN algorithm. According to the results, the proposed method was able to estimate the BGC with a Mean Absolute Error (MAE) of 0.0539, and a Mean Squared Error (MSE) of 0.1604. In addition, the linear relationship between BGC and ECG features has been confirmed in this paper.

Conclusion: In this paper, we propose the potential use of ECG features to determine the BGC. Additionally, we confirmed the linear relationship between BGC and ECG features. That fact will open new perspectives for further research, namely physiological models. Furthermore, the findings point to the possible application of ECG wearable devices for non-invasive continuous blood glucose monitoring via machine learning.

Advanced Bioelectrical Signal Processing Methods: Past, Present and Future Approach-Part I: Cardiac Signals

Advanced signal processing methods are one of the fastest developing scientific and technical areas of biomedical engineering with increasing usage in current clinical practice. This paper presents an extensive literature review of the methods for the digital signal processing of cardiac bioelectrical signals that are commonly applied in today’s clinical practice. This work covers the definition of bioelectrical signals. It also covers to the extreme extent of classical and advanced approaches to the alleviation of noise contamination such as digital adaptive and non-adaptive filtering, signal decomposition methods based on blind source separation and wavelet transform.

Precision Medicine and Artificial Intelligence: A Pilot Study on Deep Learning for Hypoglycemic Events Detection based on ECG

Tracking the fluctuations in blood glucose levels is important for healthy subjects and crucial diabetic patients. Tight glucose monitoring reduces the risk of hypoglycemia, which can result in a series of complications, especially in diabetic patients, such as confusion, irritability, seizure and can even be fatal in specific conditions. Hypoglycemia affects the electrophysiology of the heart. However, due to strong inter-subject heterogeneity, previous studies based on a cohort of subjects failed to deploy electrocardiogram (ECG)-based hypoglycemic detection systems reliably. The current study used personalised medicine approach and Artificial Intelligence (AI) to automatically detect nocturnal hypoglycemia using a few heartbeats of raw ECG signal recorded with non-invasive, wearable devices, in healthy individuals, monitored 24 hours for 14 consecutive days. Additionally, we present a visualisation method enabling clinicians to visualise which part of the ECG signal (e.g., T-wave, ST-interval) is significantly associated with the hypoglycemic event in each subject, overcoming the intelligibility problem of deep-learning methods. These results advance the feasibility of a real-time, non-invasive hypoglycemia alarming system using short excerpts of ECG signal.

Principal component model for maternal ECG extraction in fetal QRS detection

Fetal cardiac monitoring using noninvasive abdominal leads can provide important information on fetal well-being. However, extraction of fetal electrocardiogram (fECG) from abdominal signals is often problematic because of the higher amplitude maternal ECG (mECG). The aim of this study was to introduce a principal component model for removing the maternal ECG from abdominal signals. The proposed method removes mECG waveforms with high accuracy even though abdominal movements cause morphological deviation to mECG complexes. The method can be used for single or multi-lead measurements. The proposed method was tested using 175 1 min long abdominal measurements with fetal QRS annotation markers acquired from a fetal scalp electrode. Using the presented mECG removing algorithm and matched filtering based fQRS detector, 95% sensitivity for fQRS detection and 4.84 ms RMS error for fetal RR-interval estimation were acquired.

Hypoglycemia detection based on cardiac repolarization features

Hypoglycemia is known to affect repolarization characteristics of the heart. These changes are shown from ECG by prolonged QT-time and T-wave flattening. In this study we constructed a classifier based on these ECG parameters. By using the classifier we tried to detect hypoglycemic events from measurements of 22 test subjects. Hypoglycemic state was achieved using glucose clamp technique. Used test protocol consisted of three stages: normoglycemic period, transition period (blood glucose concentration decreasing) and hypoglycemic period. Subjects were divided into three groups: 9 healthy controls (Healthy), 6 otherwise healthy type 1 diabetics (T1DM) and 7 type 1 diabetics with disease complications (T1DMc). Detection of hypoglycemic event could be made passably from 15/22 measurements. In addition, we found that detection process is easier for healthy and T1DM groups than T1DMc group diabetics because in T1DMc group subjects' have lower autonomic response to hypoglycemic events. Also we noticed that changes in ECG occurs few minutes after blood glucose is decreased below 3.5 mmol/1.

Dynamic estimation of cardiac repolarization characteristics during hypoglycemia in healthy and diabetic subjects

Hypoglycemia is known to affect the repolarization characteristics of the heart, but the mechanisms behind these changes are not completely understood. We analyzed repolarization characteristics continuously from 22 subjects during normoglycemic period, transition period (blood glucose concentration decreasing) and hypoglycemic period from nine healthy controls (Healthy), six otherwise healthy type 1 diabetics (T1DM) and seven type 1 diabetics with disease complications (T1DMc). An advanced principal component regression (PCR)-based method was used for estimating ECG parameters beat-by-beat, and thus, continuous comparison between the repolarization characteristics and blood glucose values was made. We observed that hypoglycemia related ECG changes in the T1DMc group were smaller than changes in the Healthy and T1DM groups. We also noticed that when glucose concentration remained at a low level, the heart rate corrected QT interval prolonged progressively. Finally, a few minutes time lag was observed between the start of hypoglycemia and cardiac repolarization changes. One explanation for these observations could be that hypoglycemia related hormonal changes have a significant role behind the repolarization changes. This could explain at least the observed time lag (hormonal changes are slow) and the lower repolarization changes in the T1DMc group (hormonal secretion lowered in long duration diabetics).