Abstract Normal pregnancy is associated with insulin resistance although the mechanism is not understood. Increased intramyocellular lipid is closely associated with the insulin resistance of type 2 diabetes and obesity, and the aim of this study was to determine whether this was so for the physiological insulin resistance of pregnancy. Eleven primiparous healthy pregnant women (age: 27–39 years, body mass index 24.0±3.1 kg/m2) and no personal or family history of diabetes underwent magnetic resonance studies to quantify intramyocellular lipid, plasma lipid fractions, and insulin sensitivity. The meal-related insulin sensitivity index was considerably lower in pregnancy (45.6±9.9 vs. 193.0±26.1; 10−4 dl/kg/min per pmol/l, p=0.0002). Fasting plasma triglyceride levels were elevated 3-fold during pregnancy (2.3±0.2 vs. 0.8±0.1 mmol/l, p<0.01) and the low-density density lipoprotein fraction, responsible for fatty acid delivery to muscle and other tissues, was 6-fold elevated (0.75±0.43 vs. 0.12±0.09 mmol/l; p=0.001). However, mean intramyocellular lipid concentrations of the soleus muscle were not different during pregnancy (20.0±2.3 vs. 19.1±3.2 mmol/l, p=0.64). The pregnancy effect on muscle insulin resistance is distinct from that underlying type 2 diabetes.

The most common automated blood pressure devices determine systolic, diastolic and mean arterial blood pressure (SBP, DBP and MAP) by analysing the oscillometric pulse waveform. The aim of this study was to assess the variability of the amplitude of oscillometric pulse waveform characteristics at SBP, DBP and MAP.

The aim was to develop a fully automatic QT interval measurement algorithm for the 2006 PhysioNet/ Computers in Cardiology Challenge. The algorithm determined the Q onset and T offset points from the average beat in each lead. The QT interval measurement was calculated from the median Q onset and T offset points. Manual measurements were also made. The mean (sd) difference between automatic and reference measurements was -49 (38) ms and between manual and reference measurements was -6 (48) ms.

INTRODUCTION:A simulator has been developed that enables previously recorded clinical oscillometric waveforms to be regenerated for testing oscillometric non-invasive blood pressure measurement devices. Two non-invasive blood pressure devices were evaluated using the simulator with its database of 243 waveforms, to assess the value of a simulator for such evaluations. METHODS:Two oscillometric non-invasive blood pressure devices, both of which had previously been validated against auscultatory references, were selected. The Omron HEM-907 (Omron, Hoofddorp, The Netherlands) measures the pressure during linear cuff deflation and the GE ProCare 400 (GE Healthcare, Tampa, Florida, USA) measures during step deflation. Each non-invasive blood pressure device was attached to the simulator and pressures were recorded from all 243 waveforms. The differences between the systolic and diastolic pressures measured by each non-invasive blood pressure device and the auscultatory references for each waveform were calculated. These were assessed with the European and American validation standards and with the British Hypertension Society protocol. RESULTS:The paired pressure differences (non-invasive blood pressure device minus auscultatory reference) for each device complied partly, but not fully, with the standards or protocol. The means (+/-standard deviation) of the paired systolic and diastolic pressures differences for the Omron were -2.4 mmHg (+/-5.9 mmHg) and -8.9 mmHg (+/-6.5 mmHg), and for the ProCare were -6.5 mmHg (+/-10.4 mmHg) and -2.9 mmHg (+/-7.0 mmHg), respectively. The pressures recorded by the Omron device met the standards for systolic pressures but failed for diastolic pressures and conversely for the ProCare. CONCLUSION:This represents the first evaluation of non-invasive blood pressure devices with a simulator that generates previously recorded clinical oscillometric waveforms. It allowed data from over 100 study participants to be used. Both devices had been previously clinically validated, but their evaluation using the simulator with its regenerated waveforms only partly met the required criteria. Although the results did not fully match previous clinical validations, these initial results give encouragement that a simulator with sufficient stored waveforms might be able to replace the difficult and expensive clinical evaluation of non-invasive blood pressure devices that has prevented many devices from being fully evaluated.

Recent studies have indicated that multichannel magnetocardiograms (MCGs), which non-invasively measure cardiac magnetic field strength from many sites above the body surface, may provide independent information to ECGs about QT dispersion. In this study QT dispersion measurement was investigated in 61-channel MCGs and 12-lead ECGs, recorded simultaneously using the Bochum multichannel SQUID system from 20 healthy volunteers. Two automatic methods for QT interval measurement were used to determine T wave end. QT dispersion, expressed as the QT interval range, was calculated across the 12 ECG leads and 61 MCG channels for each subject. MCG channels were then systematically excluded in random groups of 10 and QT dispersion was calculated for 51, 41, 31, 21 and 11 channels respectively. Dispersion in the 61 channel MCG was significantly greater than 12 lead ECG by 39.1 (20.4) ms (mean (SD)) (p<0.00001), across techniques and all subjects. Significant differences of 34.7 (18.7) ms (p<0.00001), 29.6 (18.8) ms (p<0.00001), 25.5 (21.3) ms (p<0.0002) and 14.9 (21.0) ms (p<0.001) were also obtained for the 51 to 21 channel data respectively. No significant differences were obtained for 11-channel data. Automatic MCG dispersion measurements were significantly greater than dispersion from ECGs even when the number of MCG channels were reduced to 21 channels. However, small numbers of MCG channels need to be selected appropriately. The results suggest that a limited number of MCG channels is of value for repolarisation dispersion

Current RR time series simulations are distinguishable from real data by automatic algorithms. We hypothesised that RR time series simulations could be improved by using time series data from naturally occurring phenomena. 20 records of annual river flow data for the river Tyne in north eastern England were obtained. Each river flow data record was used to generate a single 24 h simulated RR time series with the property of self similarity. We compared the standard frequency parameters ULF, VLF, LF and HF normalised to the total power, for the simulated RR, with those from physiological data from 20 subjects. The river flow data produced realistic simulations of RR time series with significant differences between physiological and simulated series for VLF only. Time series data from river flow or other naturally occurring phenomena may provide useful components in producing RR time series with more realistic characteristics than current artificially generated data

We assessed characteristics of atrial and ventricular activity from the ECG for predicting the oflset of atrial fibrillation for the 2004 PhysioNet/Computers in Cardiology Challenge. Seven parameters were analysed with five based on the statistical characteristics of the RR interval (mean, standard deviation, skewness, kurtosis and median beat-to-beat change), and fibrillation frequency and atrial signal amplitude. The power of the parameters to predict termination of the arrhythmia was assessed individually and in combination using linear discriminant analysis ( D A ) and a?t$cial neural network (ANN) techniques. Fibrillation fiequency with a threshold value of 5.55 Hz was able to idenrify 10/10 learning set records which terminated immediately (T) and 8/10 of non-terminating records (N) and was the best of the individual parameters. Classifcations for the test set for event 1 of the challenge for algorithms based on fibrillation frequency alone, LDA and ANN received scores of 24/30, 18/30 and 23/30 respectively. Low jibrillation frequency is an indicator of spontaneous termination of atrial fibrillation.

Multichannel magnetocardiograms (MCGs) and 12- Lead ECGs were recorded simultuneousiy using the PTB mulrickannel SQUID system from 8 normal volunteers. An automatic method for QT interval measurement was used to determine T wave end with measurements made over 10 consecutive beats in both the MCG and ECG. Channeldeads with small nbsoiute T wave amplitudes and smallest and longest QT measuremenrs were excluded from the analysis. QT dispersion, expressed as the QT interval range, was calculated for a single beat and the average QT interval for each channemead over IO beats. Mean QT dispersion measurement (standard deviation) for the single beat was 44 (26) ms for MCG and 37 {35) ms for ECG over all subjects. Averaging over IO beats reduced mean QT dispersion to 36.1 (14.6) ms for MCG and 20.9 (13.2) ms for ECG. QT dispersion in the MCG was greater than in the ECG, by 7 ms for the single beat and by 15 m (p < 0.03) for averaged datu over all subjects. Averaging influenced ventricular dispersion measurements in both MCG and ECG waveforms. There were differences in dispersion of ventricular repolarisation time between ECG and MCG, with MCG significantly greater than ECG for averaged data.

Multichannel MCG noninvasively measures cardiac magnetic field strength from many sites at the body surface, potentially providing useful regional information about ventricular repolarization. Previous work on ECGs has shown that automatic techniques for repolarization measurement are better than manual measurement at discriminating patients with cardiac conditions from normal subjects. Although automatic repolarization measurement techniques have been quantified for ECGs, no comparative data exists for the MCG. In this study four different automatic repolarization (QT) interval techniques for detecting T wave end in the MCG were compared. The influence of MCG filtering on the automatic algorithms was also quantified. MCGs were obtained at 49 sites over the heart from 23 normal subjects. Automatic measurements of the repolarization (QT) interval were made following the addition of different high pass (0.25, 0.5, 1 Hz) and low pass (100, 60, 40, 30 Hz) filters. There were consistent differences between automatic techniques in the unfiltered data amounting to greatest mean difference of 52.3 ms. Low pass filtering significantly increased the automatic repolarization (QT) interval relative to unfiltered measurement by 6.5 (3.2) ms (mean SD) for 100 Hz, 6.0 (3.0) ms for 60 Hz, 8.1 (3.2) ms for 40 Hz, and 8.8 (3.1) ms for 30 Hz across all techniques. High pass filtering significantly decreased the value by −2.6 (6.0) ms for 0.25 Hz, −5.5 (5.3) ms for 0.5 Hz, and −17.1 (7.8) ms for 1 Hz. Automatic measurements of repolarization (QT) in the MCG differ between techniques and are influenced by filtering. These effects should be considered when comparing results. (PACE 2003; 26:2096–2102)

Multichannel magnetocardiography measures the magnetic field distribution of the human heart noninvasively from many sites over the body surface. Multichannel magnetocardiogram (MCG) analysis enables regional temporal differences in the distribution of cardiac magnetic field strength during depolarization and repolarization to be identified, allowing estimation of the global and local inhomogeneity of the cardiac activation process. The aim of this study was to compare the spatial distribution of cardiac magnetic field strength during ventricular depolarization and repolarization in both normal subjects and patients with cardiac abnormalities, obtaining amplitude measurements by magnetocardiography. MCGs were recorded at 49 sites over the heart from three normal subjects and two patients with inverted T-wave conditions. The magnetic field intensity during depolarization and repolarization was measured automatically for each channel and displayed spatially as contour maps. A Pearson correlation was used to determine the spatial relationship between the variables. For normal subjects, magnetic field strength maps during depolarization (R-wave) showed two asymmetric regions of magnetic field strength with a high positive value in the lower half of the chest and a high negative value above this. The regions of high R-wave amplitude corresponded spatially to concentrated asymmetric regions of high magnetic field strength during repolarization (T-wave). Pearson-r correlation coefficients of 0.7 (p<0.01), 0.8 (p<0.01) and 0.9 (p<0.01) were obtained from this analysis for the three normal subjects. A negative correlation coefficient of -0.7 (p<0.01) was obtained for one of the subjects with inverted T-wave abnormalities, suggesting similar but inverted magnetic field and current distributions to normal subjects. Even with the high correlation values in these four subjects, the MCG was able to identify differences in the distribution of magnetic field strength, with a shift in the T-wave relative to the R-wave. The measurement of cardiac magnetic field distribution during depolarization and repolarization of normal subjects and patients with clinical abnormalities should enable the improvement of theoretical models for the explanation of the cardiac depolarization and repolarization processes.

Multichannel magnetocardiography (MCG) noninvasively measures variations in magnetic field strength from many sites at the body surface, potentially providing useful regional information about ventricular repolarization. MCGs contain features similar to ECGs, and although errors associated with repolarization measurement have been quantified for ECGs, no comparative data exists for MCGs. In this study, errors in manual measurement of repolarization interval in the MCG were determined. Sixteen MCG channels and three ECG leads were recorded simultaneously in eight healthy subjects. Each recording was displayed in a random order on a computer screen, in presentations with different noise levels, time display widths, and amplitude display heights. In total, manual measurement of repolarization intervals in 2,048 (eight subjects x 16 channels x eight presentations x two repeats) MCGs were made by each of four analysts. Measured repolarization intervals were reduced by 3 ms when noise was added and by a further 3 ms when this noise was doubled. Intervals were shortened by 9 ms when the time display width was doubled and by a further 10 ms when the display width was doubled again. Measurements increased by 7 ms for a doubling of amplitude display height, equivalent to a doubling of T wave height. There were also consistent differences between analysts; amounting to a greatest mean difference of 24 ms. Display characteristics, added noise, and different analysts thus affect manual repolarization interval measurements in MCG. The errors detected demonstrate the importance of a standard presentation for repolarization measurement in the MCG.

Cardiac repolarisation can be detected by magnetocardiograms (MCGs) which non-invasively measure the variation in magnetic field strength at the body surface. The aim of this study was to assess quantitatively the influence of MCG filtering on repolarisation interval measurements. A technique for automatic analysis of ECG signals, which used modelling of the terminal T wave section to determine the end point, was extended and applied to multichannel MCG recordings of 8 normal subjects. Automatic repolarisation interval measurements were made following the addition of high and low pass filters. An experienced analyst also manually measured repolarisation intervals of the unfiltered data. The automatic technique underestimated repolarisation interval in the unfiltered data relative to manual measurement by 34.1 (8.9) ms (mean (standard deviation)). Low pass, 40Hz, filtering increased repolarisation intervals relative to unfiltered measurements by 5.1 (1.3) ms. High pass, 0.5 Hz, filtering decreased the values by 7.6 (5.0) ms.

An RR interval simulator was developed as part of the Computers in Cardiology Challenge (entry no. 184). The simulator was based on observed physiological changes in normal subjects. A template, which represented slow trends in RR interval during 24 hours, was derived from a number of parameters that represent sleep and wake states. The variations about the template were generated so that the frequency spectrum of the final signal was similar to data from normal subjects. The frequency spectrum of normal RR intervals shows a strong 1/f component, a peak at around 0.1 Hz and a peak corresponding to respiratory rate between 0.15 Hz and 0.4 Hz. These were simulated by adding to the template pink noise, a number of random phased sinusoids at around 0.1 Hz and a signal that represented respiration. Variations due to respiration were dependent on sleep and wake states.

Fifty sequences of PhysioNet R-to-R interval data, covering periods of between 20 and 24 hours, were classified into real or simulated groups. The RR interval characteristics were investigated in both the time domain and frequency domain. Eleven characteristics were analysed, and the range of measurements for each was studied for outliers from the main distribution. In the time domain, a restricted pattern of RR interval distributions classified 4 sequences as abnormal, and a reduced RR variability produced 18 classifications, with an overlap of 8, giving a total of 14/50 as abnormal. In the frequency domain, abnormally restricted very low frequency characteristics produced 26 classifications as abnormal with 10 overlaps giving a total of 16. The low frequency to high frequency ratio classified 4 as abnormal, but three of these were already detected by abnormal low frequency characteristics, giving a total of 17 classified in the frequency domain. Of the 17 classified in the frequency domain and of the 14 in the time domain there was an overlap of 9, resulting in 22 abnormal classifications, and suggesting that these were simulated. When PhysioNet assessed this classification a correct grouping of 100% was achieved on a single entry (reference 20020426.082234).

Atrial fibrillation is an ECG rhythm with a significant mortality due to stroke. The objective of this study was to detect those patients most likely to develop atrial fibrillation, and to identify ECGs closest to the onset of fibrillation. Our hypothesis was that patients with atrial fibrillation would have atrial ectopy, and the frequency of this activity would increase prior to onset of fibrillation. From a learning set of 100 30-minute ECGs from 50 patients, 25 without atrial fibrillation (normal) and 25 who subsequently developed atrial fibrillation, an algorithm was developed to detect the presence of ectopic beats using R-R interval data. In the learning set, 37/50 abnormal and 34/50 normal patients were identified, giving a potential screening accuracy of 71%. As a prediction test to detect the ECGs closest to atrial fibrillation, 19/25 were correctly identified. For the test set, a total of 29/50 were correctly assigned to the normal and fibrillation groups, and a 39/50 score obtained in predicting the onset of atrial fibrillation

Spatial distribution of repolarisation interval and T-wave amplitude were examined using multi-channel magnetocardiograms (MCGs), which non-invasively measure variations in magnetic field strength at the body surface. MCGs were obtained using the Physikalisch-Technische Bundesanstalt (PTB) multi-channel SQUID system at 49 sites over the heart from eight healthy volunteers. Each recording channel was displayed on a computer screen. Repolarisation intervals and T-wave amplitudes in each channel were measured manually and displayed graphically as contour maps. Pearson correlation was used to determine the spatial relationship between the two variables. Mean(standard deviation) values of repolarisation interval and T-wave amplitude were 381(7) ms and 4(2) pT respectively. The standard deviation of the difference between repeat measurements for repolarisation interval and T-wave amplitude was 5.2 ms and 0.4 pT respectively. Repolarisation measurements tended to increase by approximately 8 ms for a doubling of T-wave height. Contour maps showed two distinct regions of high repolarisation interval (>380 ms), corresponding spatially to concentrated areas of high absolute T-wave amplitude (>4 pT). A Pearson correlation coefficient of 0.7 (p<0.0005) was obtained from this analysis. Repolarisation intervals and T-wave heights in normal subjects recorded from magnetocardiography have distinctive but related spatial distributions