Delineation of 12-Lead ECG Representative Beats Using Convolutional Encoder-Decoders with Residual and Recurrent Connections.

The aim of this study is to address the challenge of 12-lead ECG delineation by different encoder-decoder architectures of deep neural networks (DNNs). This study compares four concepts for encoder-decoders based on a fully convolutional architecture (CED-Net) and its modifications with a recurrent layer (CED-LSTM-Net), residual connections between symmetrical encoder and decoder feature maps (CED-U-Net), and sequential residual blocks (CED-Res-Net). All DNNs transform 12-lead representative beats to three diagnostic ECG intervals (P-wave, QRS-complex, QT-interval) used for the global delineation of the representative beat (P-onset, P-offset, QRS-onset, QRS-offset, T-offset). All DNNs were trained and optimized using the large PhysioNet ECG database (PTB-XL) under identical conditions, applying an advanced approach for machine-based supervised learning with a reference algorithm for ECG delineation (ETM, Schiller AG, Baar, Switzerland). The test results indicate that all DNN architectures are equally capable of reproducing the reference delineation algorithm's measurements in the diagnostic PTB database with an average P-wave detection accuracy (96.6%) and time and duration errors: mean values (-2.6 to 2.4 ms) and standard deviations (2.9 to 11.4 ms). The validation according to the standard-based evaluation practices of diagnostic electrocardiographs with the CSE database outlines a CED-Net model, which measures P-duration (2.6 ± 11.0 ms), PQ-interval (0.9 ± 5.8 ms), QRS-duration (-2.4 ± 5.4 ms), and QT-interval (-0.7 ± 10.3 ms), which meet all standard tolerances. Noise tests with high-frequency, low-frequency, and power-line frequency noise (50/60 Hz) confirm that CED-Net, CED-Res-Net, and CED-LSTM-Net are robust to all types of noise, mostly presenting a mean duration error < 2.5 ms when compared to measurements without noise. Reduced noise immunity is observed for the U-net architecture. Comparative analysis with other published studies scores this research within the lower range of time errors, highlighting its competitive performance.

Simulating Arbitrary Electrode Reversals in Standard 12-lead ECG.

Electrode reversal errors in standard 12-lead electrocardiograms (ECG) can produce significant ECG changes and, in turn, misleading diagnoses. Their detection is important but mostly limited to the design of criteria using ECG databases with simulated reversals, without Wilson's central terminal (WCT) potential change. This is, to the best of our knowledge, the first study that presents an algebraic transformation for simulation of all possible ECG cable reversals, including those with displaced WCT, where most of the leads appear with distorted morphology. The simulation model of ECG electrode swaps and the resultant WCT potential change is derived in the standard 12-lead ECG setup. The transformation formulas are theoretically compared to known limb lead reversals and experimentally proven for unknown limb-chest electrode swaps using a 12-lead ECG database from 25 healthy volunteers (recordings without electrode swaps and with 5 unicolor pairs swaps, including red (right arm-C1), yellow (left arm-C2), green (left leg (LL) -C3), black (right leg (RL)-C5), all unicolor pairs). Two applications of the transformation are shown to be feasible: 'Forward' (simulation of reordered leads from correct leads) and 'Inverse' (reconstruction of correct leads from an ECG recorded with known electrode reversals). Deficiencies are found only when the ground RL electrode is swapped as this case requires guessing the unknown RL electrode potential. We suggest assuming that potential to be equal to that of the LL electrode. The 'Forward' transformation is important for comprehensive training platforms of humans and machines to reliably recognize simulated electrode swaps using the available resources of correctly recorded ECG databases. The 'Inverse' transformation can save time and costs for repeated ECG recordings by reconstructing the correct lead set if a lead swap is detected after the end of the recording. In cases when the electrode reversal is unknown but a prior correct ECG recording of the same patient is available, the 'Inverse' transformation is tested to detect the exact swapping of the electrodes with an accuracy of (96% to 100%).

Ranking of the most reliable beat morphology and heart rate variability features for the detection of atrial fibrillation in short single-lead ECG.

OBJECTIVE: This study participated in the 2017 PhysioNet/CinC Challenge dedicated to the classification of atrial fibrillation (AF), normal sinus rhythm (Normal), other arrhythmia (Other) and strong noise, using single-lead electrocardiogram (ECG) recordings with a duration <60 s. The aim is to apply a linear threshold-based strategy for arrhythmia classification, ranking the most powerful time domain ECG features that could be easily reproduced on any platform. APPROACH: An algorithm for time domain ECG analysis was designed to extract 44 features with focus on the following: noise detection; heart rate variability (HRV) analysis; beat morphology analysis and delineation of P-, QRS-, and T-waves in the robust average beat; detection of atrial activity by the presence of P-waves in the average beat and atrial fibrillatory waves (f-waves) during TQ intervals. A linear discriminant analysis (LDA) classifier was optimized on the Challenge training set (8528 ECGs) by stepwise selection of a nonredundant feature set until maximization of the Challenge F1 score. Heart rate (HR) was an independent factor for the LDA classifier design, particular to bradycardia (HR ⩽ 50 bpm), normal rhythm (HR = 50-100 bpm), tachycardia (HR ⩾ 100 bpm). MAIN RESULTS: The algorithm obtained official Challenge F1 scores of 0.80 (Overall), 0.90 (Normal), 0.81 (AF), 0.70 (Other), and 0.54 (Noise) on the hidden Challenge test set (3658 ECGs). This is equivalent to a true positive rate (TPR) = 90.1% (Normal), 81.5% (AF), 67.7% (Other), and 69.5% (Noise), and a false positive rate (FPR) = 13.6% (Normal), 2.3% (AF), 7.7% (Other), and 1.5% (Noise). SIGNIFICANCE: The top five features, which together contributed to about 94% of the maximal F1 score were ranked: (1) proportion of RR intervals differing by >50 ms from the preceding RR interval; (2) Poincaré plot geometry estimated by the ratio of the minor-to-major semi-axes of the fitted ellipse; (3) P-wave presence in the average beat; (4) mean percentage of the RR interval first differences; and (5) mean correlation of all beats against the average beat. The global rank of feature extraction methods highlighted that HRV alone was able to provide 92.5% of the maximal F1 score (0.74 versus 0.8). The added value of more complex ECG morphology analysis was less significant for Normal, AF, and Other rhythms (+0.02 to 0.08 points) than for Noise (+0.19 points); however, these were indispensable for wearable ECG recording devices with frequent artefact disturbance.

Comparison of automated interval measurements by widely used algorithms in digital electrocardiographs.

BACKGROUND: Automated measurements of electrocardiographic (ECG) intervals by current-generation digital electrocardiographs are critical to computer-based ECG diagnostic statements, to serial comparison of ECGs, and to epidemiological studies of ECG findings in populations. A previous study demonstrated generally small but often significant systematic differences among 4 algorithms widely used for automated ECG in the United States and that measurement differences could be related to the degree of abnormality of the underlying tracing. Since that publication, some algorithms have been adjusted, whereas other large manufacturers of automated ECGs have asked to participate in an extension of this comparison. METHODS: Seven widely used automated algorithms for computer-based interpretation participated in this blinded study of 800 digitized ECGs provided by the Cardiac Safety Research Consortium. All tracings were different from the study of 4 algorithms reported in 2014, and the selected population was heavily weighted toward groups with known effects on the QT interval: included were 200 normal subjects, 200 normal subjects receiving moxifloxacin as part of an active control arm of thorough QT studies, 200 subjects with genetically proved long QT syndrome type 1 (LQT1), and 200 subjects with genetically proved long QT syndrome Type 2 (LQT2). RESULTS: For the entire population of 800 subjects, pairwise differences between algorithms for each mean interval value were clinically small, even where statistically significant, ranging from 0.2 to 3.6milliseconds for the PR interval, 0.1 to 8.1milliseconds for QRS duration, and 0.1 to 9.3milliseconds for QT interval. The mean value of all paired differences among algorithms was higher in the long QT groups than in normals for both QRS duration and QT intervals. Differences in mean QRS duration ranged from 0.2 to 13.3milliseconds in the LQT1 subjects and from 0.2 to 11.0milliseconds in the LQT2 subjects. Differences in measured QT duration (not corrected for heart rate) ranged from 0.2 to 10.5milliseconds in the LQT1 subjects and from 0.9 to 12.8milliseconds in the LQT2 subjects. CONCLUSIONS: Among current-generation computer-based electrocardiographs, clinically small but statistically significant differences exist between ECG interval measurements by individual algorithms. Measurement differences between algorithms for QRS duration and for QT interval are larger in long QT interval subjects than in normal subjects. Comparisons of population study norms should be aware of small systematic differences in interval measurements due to different algorithm methodologies, within-individual interval measurement comparisons should use comparable methods, and further attempts to harmonize interval measurement methodologies are warranted.

Perspectives of human verification via binary QRS template matching of single-lead and 12-lead electrocardiogram.

OBJECTIVE: This study aims to validate the 12-lead electrocardiogram (ECG) as a biometric modality based on two straightforward binary QRS template matching characteristics. Different perspectives of the human verification problem are considered, regarding the optimal lead selection and stability over sample size, gender, age, heart rate (HR). METHODS: A clinical 12-lead resting ECG database, including a population of 460 subjects with two-session recordings (>1 year apart) is used. Cost-effective strategies for extraction of personalized QRS patterns (100ms) and binary template matching estimate similarity in the time scale (matching time) and dissimilarity in the amplitude scale (mismatch area). The two-class person verification task, taking the decision to validate or to reject the subject identity is managed by linear discriminant analysis (LDA). Non-redundant LDA models for different lead configurations (I,II,III,aVF,aVL,aVF,V1-V6) are trained on the first half of 230 subjects by stepwise feature selection until maximization of the area under the receiver operating characteristic curve (ROC AUC). The operating point on the training ROC at equal error rate (EER) is tested on the independent dataset (second half of 230 subjects) to report unbiased validation of test-ROC AUC and true verification rate (TVR = 100-EER). The test results are further evaluated in groups by sample size, gender, age, HR. RESULTS AND DISCUSSION: The optimal QRS pattern projection for single-lead ECG biometric modality is found in the frontal plane sector (60°-0°) with best (Test-AUC/TVR) for lead II (0.941/86.8%) and slight accuracy drop for -aVR (-0.017/-1.4%), I (-0.01/-1.5%). Chest ECG leads have degrading accuracy from V1 (0.885/80.6%) to V6 (0.799/71.8%). The multi-lead ECG improves verification: 6-chest (0.97/90.9%), 6-limb (0.986/94.3%), 12-leads (0.995/97.5%). The QRS pattern matching model shows stable performance for verification of 10 to 230 individuals; insignificant degradation of TVR in women by (1.2-3.6%), adults ≥70 years (3.7%), younger <40 years (1.9%), HR<60bpm (1.2%), HR>90bpm (3.9%), no degradation for HR change (0 to >20bpm).

Human Identification by Cross-Correlation and Pattern Matching of Personalized Heartbeat: Influence of ECG Leads and Reference Database Size.

Human identification (ID) is a biometric task, comparing single input sample to many stored templates to identify an individual in a reference database. This paper aims to present the perspectives of personalized heartbeat pattern for reliable ECG-based identification. The investigations are using a database with 460 pairs of 12-lead resting electrocardiograms (ECG) with 10-s durations recorded at time-instants T1 and T2 > T1 + 1 year. Intra-subject long-term ECG stability and inter-subject variability of personalized PQRST (500 ms) and QRS (100 ms) patterns is quantified via cross-correlation, amplitude ratio and pattern matching between T1 and T2 using 7 features × 12-leads. Single and multi-lead ID models are trained on the first 230 ECG pairs. Their validation on 10, 20, ... 230 reference subjects (RS) from the remaining 230 ECG pairs shows: (i) two best single-lead ID models using lead II for a small population RS = (10-140) with identification accuracy AccID = (89.4-67.2)% and aVF for a large population RS = (140-230) with AccID = (67.2-63.9)%; (ii) better performance of the 6-lead limb vs. the 6-lead chest ID model-(91.4-76.1)% vs. (90.9-70)% for RS = (10-230); (iii) best performance of the 12-lead ID model-(98.4-87.4)% for RS = (10-230). The tolerable reference database size, keeping AccID > 80%, is RS = 30 in the single-lead ID scenario (II); RS = 50 (6 chest leads); RS = 100 (6 limb leads), RS > 230-maximal population in this study (12-lead ECG).

Quantification of the first-order high-pass filter's influence on the automatic measurements of the electrocardiogram.

BACKGROUND AND OBJECTIVE: The first-order high-pass filter (AC coupling) has previously been shown to affect the ECG for higher cut-off frequencies. We seek to find a systematic deviation in computer measurements of the electrocardiogram when the AC coupling with a 0.05 Hz first-order high-pass filter is used. METHODS: The standard 12-lead electrocardiogram from 1248 patients and the automated measurements of their DC and AC coupled version were used. We expect a large unipolar QRS-complex to produce a deviation in the opposite direction in the ST-segment. RESULTS: We found a strong correlation between the QRS integral and the offset throughout the ST-segment. The coefficient for J amplitude deviation was found to be -0.277 µV/(µV⋅s). CONCLUSIONS: Potential dangerous alterations to the diagnostically important ST-segment were found. Medical professionals and software developers for electrocardiogram interpretation programs should be aware of such high-pass filter effects since they could be misinterpreted as pathophysiology or some pathophysiology could be masked by these effects.

Diagnostic and Prognostic Value of Lead aVR During Exercise Testing in Patients Suspected of Having Myocardial Ischemia.

We aimed to assess the diagnostic and prognostic value of ST-segment deviation in aVR, a lead often ignored in clinical practice, during exercise testing and to compare it to the most widely used criterion of ST-segment depression in V5. We enrolled 1,596 patients with suspected myocardial ischemia referred for nuclear perfusion imaging undergoing bicycle stress testing. ST-segment amplitudes in leads aVR and V5 were automatically measured. The presence of inducible myocardial ischemia was the diagnostic end point and adjudicated based on nuclear perfusion imaging and coronary angiography. Major adverse cardiac events (MACE) during 2 years of follow-up including death, acute myocardial infarction, and coronary revascularization were the prognostic end point. Exercise-induced myocardial ischemia was detected in 470 patients (29%). Median ST amplitudes for leads aVR and V5 differed significantly among patients with and without ischemia (p <0.01). The diagnostic accuracy of ST changes for myocardial ischemia as quantified by the area under the receiver operating characteristic curve was highest 2 minutes into recovery and similar in aVR and V5 (0.62, 95% confidence interval CI 0.60 to 0.65 vs 0.60, 95% confidence interval 0.58 to 0.63, p = 0.08 for comparison). In multivariate analysis, ST changes in lead aVR, but not lead V5, contributed independent diagnostic information on top of clinical parameters and manual electrocardiographic interpretation. Within 2 years of follow-up, MACE occurred in 33% of patients with ST elevations in aVR and in 16% without (p <0.001). In conclusion, ST elevation in lead aVR during exercise testing indicates inducible myocardial ischemia independently of ST depressions in lead V5 and clinical factors and also predicts MACE during follow-up.

Diagnostic and prognostic values of the V-index, a novel ECG marker quantifying spatial heterogeneity of ventricular repolarization, in patients with symptoms suggestive of non-ST-elevation myocardial infarction.

BACKGROUND: The V-index is an ECG marker quantifying spatial heterogeneity of ventricular repolarization. We prospectively assessed the diagnostic and prognostic values of the V-index in patients with suspected non-ST-elevation myocardial infarction (NSTEMI). METHODS: We prospectively enrolled 497 patients presenting with suspected NSTEMI to the emergency department (ED). Digital 12-lead ECGs of five-minute duration were recorded at presentation. The V-index was automatically calculated in a blinded fashion. Patients with a QRS duration >120ms were ruled out from analysis. The final diagnosis was adjudicated by two independent cardiologists. The prognostic endpoint was all-cause mortality during 24months of follow-up. RESULTS: NSTEMI was the final diagnosis in 14% of patients. V-index levels were higher in patients with AMI compared to other causes of chest pain (median 23ms vs. 18ms, p<0.001). The use of the V-index in addition to conventional ECG-criteria improved the diagnostic accuracy for the diagnosis of NSTEMI as quantified by area under the ROC curve from 0.66 to 0.73 (p=0.001) and the sensitivity of the ECG for AMI from 41% to 86% (p<0.001). Cumulative 24-month mortality rates were 99.4%, 98.4% and 88.3% according to tertiles of the V-index (p<0.001). After adjustment for age and important ECG and clinical parameters, the V-index remained an independent predictor of death. CONCLUSIONS: The V-index, an ECG marker quantifying spatial heterogeneity of ventricular repolarization, significantly improves the accuracy and sensitivity of the ECG for the diagnosis of NSTEMI and independently predicts mortality during follow-up.

Pseudo-real-time low-pass filter in ECG, self-adjustable to the frequency spectra of the waves.

The electrocardiogram (ECG) acquisition is often accompanied by high-frequency electromyographic (EMG) noise. The noise is difficult to be filtered, due to considerable overlapping of its frequency spectrum to the frequency spectrum of the ECG. Today, filters must conform to the new guidelines (2007) for low-pass filtering in ECG with cutoffs of 150 Hz for adolescents and adults, and to 250 Hz for children. We are suggesting a pseudo-real-time low-pass filter, self-adjustable to the frequency spectra of the ECG waves. The filter is based on the approximation procedure of Savitzky-Golay with dynamic change in the cutoff frequency. The filter is implemented pseudo-real-time (real-time with a certain delay). An additional option is the automatic on/off triggering, depending on the presence/absence of EMG noise. The analysis of the proposed filter shows that the low-frequency components of the ECG (low-power P- and T-waves, PQ-, ST- and TP-segments) are filtered with a cutoff of 14 Hz, the high-power P- and T-waves are filtered with a cutoff frequency in the range of 20-30 Hz, and the high-frequency QRS complexes are filtered with cutoff frequency of higher than 100 Hz. The suggested dynamic filter satisfies the conflicting requirements for a strong suppression of EMG noise and at the same time a maximal preservation of the ECG high-frequency components.

A Correction Formula for the ST-Segment Measurements of AC-Coupled Electrocardiograms.

GOAL: The ST segment of an electrocardiogram (ECG) is very important for the correct diagnosis of an acute myocardial infarction. Most clinical ECGs are recorded using an ACcoupled ECG amplifier. It is well known, that first-order high-pass filters used for the AC coupling can affect the ST segment of an ECG. This effect is stronger the higher the filter's cut-off frequency is and the larger the QRS integral is. We present a formula that estimates these changes in the ST segment and therefore allows for correcting ST measurements that are based on an ACcoupled ECG. METHODS: The presented correction formula can be applied when only four parameters are known: the possibly estimated QRS area A, the QRS duration W, the beat-to-beat interval TRR, and the filter time constant T, further, the time point Tj to correct-after the J point-must be specified. RESULTS: The formula is correct within 0.6% until 40% ms after the J point and within 6% until 80 ms after the J point. CONCLUSION AND SIGNIFICANCE: It is not necessary to have the raw data available and the formula therefore opens up the possibility of reevaluating studies that are based on ACcoupled ECGs and compare the results of such studies with studies that are based on newer, DC-coupled ECGs.

Intersubject variability and intrasubject reproducibility of 12-lead ECG metrics: Implications for human verification.

BACKGROUND: Electrocardiogram (ECG) biometrics is an advanced technology, not yet covered by guidelines on criteria, features and leads for maximal authentication accuracy. OBJECTIVE: This study aims to define the minimal set of morphological metrics in 12-lead ECG by optimization towards high reliability and security, and validation in a person verification model across a large population. METHODS: A standard 12-lead resting ECG database from 574 non-cardiac patients with two remote recordings (>1year apart) was used. A commercial ECG analysis module (Schiller AG) measured 202 morphological features, including lead-specific amplitudes, durations, ST-metrics, and axes. Coefficient of variation (CV, intersubject variability) and percent-mean-absolute-difference (PMAD, intrasubject reproducibility) defined the optimization (PMAD/CV→min) and restriction (CV<30%) criteria for selection of the most stable and distinctive features. Linear discriminant analysis (LDA) validated the non-redundant feature set for person verification. RESULTS AND CONCLUSIONS: Maximal LDA verification sensitivity (85.3%) and specificity (86.4%) were validated for 11 optimal features: R-amplitude (I,II,V1,V2,V3,V5), S-amplitude (V1,V2), Tnegative-amplitude (aVR), and R-duration (aVF,V1).

Inter-lead correlation analysis for automated detection of cable reversals in 12/16-lead ECG.

BACKGROUND AND OBJECTIVE: A crucial factor for proper electrocardiogram (ECG) interpretation is the correct electrode placement in standard 12-lead ECG and extended 16-lead ECG for accurate diagnosis of acute myocardial infarctions. In the context of optimal patient care, we present and evaluate a new method for automated detection of reversals in peripheral and precordial (standard, right and posterior) leads, based on simple rules with inter-lead correlation dependencies. METHODS: The algorithm for analysis of cable reversals relies on scoring of inter-lead correlations estimated over 4s snapshots with time-coherent data from multiple ECG leads. Peripheral cable reversals are detected by assessment of nine correlation coefficients, comparing V6 to limb leads: (I, II, III, -I, -II, -III, -aVR, -aVL, -aVF). Precordial lead reversals are detected by analysis of the ECG pattern cross-correlation progression within lead sets (V1-V6), (V4R, V3R, V3, V4), and (V4, V5, V6, V8, V9). Disturbed progression identifies the swapped leads. RESULTS: A test-set, including 2239 ECGs from three independent sources-public 12-lead (PTB, CSE) and proprietary 16-lead (Basel University Hospital) databases-is used for algorithm validation, reporting specificity (Sp) and sensitivity (Se) as true negative and true positive detection of simulated lead swaps. Reversals of limb leads are detected with Se = 95.5-96.9% and 100% when right leg is involved in the reversal. Among all 15 possible pairwise reversals in standard precordial leads, adjacent lead reversals are detected with Se = 93.8% (V5-V6), 95.6% (V2-V3), 95.9% (V3-V4), 97.1% (V1-V2), and 97.8% (V4-V5), increasing to 97.8-99.8% for reversals of anatomically more distant electrodes. The pairwise reversals in the four extra precordial leads are detected with Se = 74.7% (right-sided V4R-V3R), 91.4% (posterior V8-V9), 93.7% (V4R-V9), and 97.7% (V4R-V8, V3R-V9, V3R-V8). Higher true negative rate is achieved with Sp > 99% (standard 12-lead ECG), 81.9% (V4R-V3R), 91.4% (V8-V9), and 100% (V4R-V9, V4R-V8, V3R-V9, V3R-V8), which is reasonable considering the low prevalence of lead swaps in clinical environment. CONCLUSIONS: Inter-lead correlation analysis is able to provide robust detection of cable reversals in standard 12-lead ECG, effectively extended to 16-lead ECG applications that have not previously been addressed.

Real-time arrhythmia detection with supplementary ECG quality and pulse wave monitoring for the reduction of false alarms in ICUs.

False intensive care unit (ICU) alarms induce stress in both patients and clinical staff and decrease the quality of care, thus significantly increasing both the hospital recovery time and rehospitalization rates. In the PhysioNet/CinC Challenge 2015 for reducing false arrhythmia alarms in ICU bedside monitor data, this paper validates the application of a real-time arrhythmia detection library (ADLib, Schiller AG) for the robust detection of five types of life-threatening arrhythmia alarms. The strength of the application is to give immediate feedback on the arrhythmia event within a scan interval of 3 s-7.5 s, and to increase the noise immunity of electrocardiogram (ECG) arrhythmia analysis by fusing its decision with supplementary ECG quality interpretation and real-time pulse wave monitoring (quality and hemodynamics) using arterial blood pressure or photoplethysmographic signals. We achieved the third-ranked real-time score (79.41) in the challenge (Event 1), however, the rank was not officially recognized due to the 'closed-source' entry. This study shows the optimization of the alarm decision module, using tunable parameters such as the scan interval, lead quality threshold, and pulse wave features, with a follow-up improvement of the real-time score (80.07). The performance (true positive rate, true negative rate) is reported in the blinded challenge test set for different arrhythmias: asystole (83%, 96%), extreme bradycardia (100%, 90%), extreme tachycardia (98%, 80%), ventricular tachycardia (84%, 82%), and ventricular fibrillation (78%, 84%). Another part of this study considers the validation of ADLib with four reference ECG databases (AHA, EDB, SVDB, MIT-BIH) according to the international recommendations for performance reports in ECG monitors (ANSI/AAMI EC57). The sensitivity (Se) and positive predictivity (+P) are: QRS detector QRS (Se, +P) > 99.7%, ventricular ectopic beat (VEB) classifier VEB (Se, +P) = 95%, and ventricular fibrillation detector VFIB (P + = 94.8%) > VFIB (Se = 86.4%), adjusted to the clinical setting requirements, giving preference to low false positive alarms.

Digital DC-Reconstruction of AC-Coupled Electrophysiological Signals with a Single Inverting Filter.

Since the introduction of digital electrocardiographs, high-pass filters have been necessary for successful analog-to-digital conversion with a reasonable amplitude resolution. On the other hand, such high-pass filters may distort the diagnostically significant ST-segment of the ECG, which can result in a misleading diagnosis. We present an inverting filter that successfully undoes the effects of a 0.05 Hz single pole high-pass filter. The inverting filter has been tested on more than 1600 clinical ECGs with one-minute durations and produces a negligible mean RMS-error of 3.1*10(-8) LSB. Alternative, less strong inverting filters have also been tested, as have different applications of the filters with respect to rounding of the signals after filtering. A design scheme for the alternative inverting filters has been suggested, based on the maximum strength of the filter. With the use of the suggested filters, it is possible to recover the original DC-coupled ECGs from AC-coupled ECGs, at least when a 0.05 Hz first order digital single pole high-pass filter is used for the AC-coupling.

Superiority of Classification Tree versus Cluster, Fuzzy and Discriminant Models in a Heartbeat Classification System.

This study presents a 2-stage heartbeat classifier of supraventricular (SVB) and ventricular (VB) beats. Stage 1 makes computationally-efficient classification of SVB-beats, using simple correlation threshold criterion for finding close match with a predominant normal (reference) beat template. The non-matched beats are next subjected to measurement of 20 basic features, tracking the beat and reference template morphology and RR-variability for subsequent refined classification in SVB or VB-class by Stage 2. Four linear classifiers are compared: cluster, fuzzy, linear discriminant analysis (LDA) and classification tree (CT), all subjected to iterative training for selection of the optimal feature space among extended 210-sized set, embodying interactive second-order effects between 20 independent features. The optimization process minimizes at equal weight the false positives in SVB-class and false negatives in VB-class. The training with European ST-T, AHA, MIT-BIH Supraventricular Arrhythmia databases found the best performance settings of all classification models: Cluster (30 features), Fuzzy (72 features), LDA (142 coefficients), CT (221 decision nodes) with top-3 best scored features: normalized current RR-interval, higher/lower frequency content ratio, beat-to-template correlation. Unbiased test-validation with MIT-BIH Arrhythmia database rates the classifiers in descending order of their specificity for SVB-class: CT (99.9%), LDA (99.6%), Cluster (99.5%), Fuzzy (99.4%); sensitivity for ventricular ectopic beats as part from VB-class (commonly reported in published beat-classification studies): CT (96.7%), Fuzzy (94.4%), LDA (94.2%), Cluster (92.4%); positive predictivity: CT (99.2%), Cluster (93.6%), LDA (93.0%), Fuzzy (92.4%). CT has superior accuracy by 0.3-6.8% points, with the advantage for easy model complexity configuration by pruning the tree consisted of easy interpretable 'if-then' rules.