Highlights of 2011

Electrical impedance tomography (EIT) is expected to become a valuable tool for monitoring mechanically ventilated patients due to its ability to continuously assess regional lung ventilation and aeration. Several sources of interference with EIT examinations exist in intensive care units (ICU). Our objectives are to demonstrate how some medical nursing and monitoring devices interfere with EIT measurements and modify the EIT scans and waveforms, which approaches can be applied to minimize these effects and how possible misinterpretation can be avoided. We present four cases of EIT examinations of adult ICU patients. Two of the patients were subjected to pulsation therapy using a pulsating air suspension mattress while being ventilated by high-frequency oscillatory or conventional pressure-controlled ventilation, respectively. The EIT signal modulation synchronous with the occurrence of the pulsating wave was 2.3 times larger than the periodic modulation synchronous with heart rate and high-frequency oscillations. During conventional ventilation, the pulsating mattress induced an EIT signal fluctuation with a magnitude corresponding to about 20% of the patient's tidal volume. In the third patient, interference with EIT examination was caused by continuous cardiac output monitoring. The last patient's examination was disturbed by impedance pneumography when excitation currents of similar frequency to EIT were used. In all subjects, the generation of functional EIT scans was compromised and interpretation of regional ventilation impossible. Discontinuation of pulsation therapy and of continuous cardiac output and impedance respiration monitoring immediately improved the EIT signal and scan quality. Offline processing of the disturbed data using frequency filtering enabled partial retrieval of relevant information. We conclude that thoracic EIT examinations in the ICU require cautious interpretation because of possible mechanical and electromagnetic interference.

The functional integrity and pathology of the skin is reflected in its electrical impedance spectra. Non-invasive electrical impedance measurements of intact skin are dominated by the high impedic stratum corneum in low frequencies and with increasing frequency gradually comes to be dominated by viable skin. Models of this multi-layered organ can increase our understanding of the actual physical properties/dimensions and facilitate better diagnostics in certain applications. Therefore, a mathematical model considering conservation of charge in the various layers of the skin and adjacent electrodes is derived and validated with experimental findings; the latter was carried out on 60 young female subjects. The impact of the stratum corneum thickness, inundation, solvent and cohort size on the electrical properties is studied. Both model parameters and experimental conditions were adjusted for calibration and subsequent validation of the model with measurements. It is found that both the model's thickness of the stratum corneum as well as experimental soaking conditions (both time and saline concentration) affect the fit between the model and measurements. It is concluded that it is essential that the electrical properties of the skin are presented in the context of the ion concentration (if a moisturizer is employed) as well as the soaking time. Further refinements should be made to determine even more accurate dielectrical properties of the stratum corneum and viable skin layers by accounting for the true skin thickness and the heterogeneity of the skin layers—this would be useful in applications where subtle alterations in the skin are of interest.

Snoring is the most common symptom of obstructive sleep apnea hypopnea syndrome (OSAHS), which is a serious disease with high community prevalence. The standard method of OSAHS diagnosis, known as polysomnography (PSG), is expensive and time consuming. There is evidence suggesting that snore-related sounds (SRS) carry sufficient information to diagnose OSAHS. In this paper we present a technique for diagnosing OSAHS based solely on snore sound analysis. The method comprises a logistic regression model fed with snore parameters derived from its features such as the pitch and total airway response (TAR) estimated using a higher order statistics (HOS)-based algorithm. Pitch represents a time domain characteristic of the airway vibrations and the TAR represents the acoustical changes brought about by the collapsing upper airways. The performance of the proposed method was evaluated using the technique of K-fold cross validation, on a clinical database consisting of overnight snoring sounds of 41 subjects. The method achieved 89.3% sensitivity with 92.3% specificity (the area under the ROC curve was 0.96). These results establish the feasibility of developing a snore-based OSAHS community-screening device, which does not require any contact measurements.

A new method to image and quantify intra-abdominal haemorrhage using electrical impedance tomography (EIT) was tested in vivo. Supine peritoneal dialysis patients were monitored using an 8-electrode hemiarray placed on the anterior abdomen. EIT measurements were recorded using the EPack II data acquisition system before, during, and after the administration of dialysate. The amount of dialysate infused was recorded synchronous with EIT measurements and used as a control. Tomographic images of impedance change were reconstructed using a weighted, sensitivity-based method and then post-processed to obtain a quantitative estimate of the total dialysate volume added and the rate of dialysate administration. Our preliminary study included two subjects, one male and one female, each of whom participated for two sessions spaced about 6 months apart. Data collected from these sessions indicated that with an in vivo SNR of about 35 dB the EPack II can detect accumulations larger than about 100 ml, with a quantification uncertainty of about 50 ml. The rate of accumulation was determined in less than 2 min. This method shows promise for automated detection of other pathologies, eg ascites, and is adaptable to detecting conductive accumulations in other anatomy.

Approximate entropy (ApEn) is widely accepted as a complexity measure of the heart rate variability (HRV) signal, but selecting the criteria for the threshold value r is controversial. This paper aims to verify whether Chon's method of forecasting the r_max is an appropriate one for the HRV signal. The standard limb lead ECG signals of 120 subjects were recorded for 10 min in a supine position. The subjects were divided into two groups: the heart failure (22 females and 38 males, median age 62.4 ± 12.6) and healthy control group (33 females and 27 males, median age 51.5 ± 16.9). Three types of ApEn were calculated: the ApEn_0.2 using the recommended constant r = 0.2, the ApEn_chon using Chon's method and the ApEn_max using the true r_max. A Wilcoxon rank sum test showed that the ApEn_0.2 (p = 0.267) and the ApEn_max (p = 0.813) had no statistical differences between the two groups, while the ApEn_chon (p = 0.040) had. We generated a synthetic database to study the effect of two influential factors (the signal length N and the ratio of short- and long-term variability sd₁/sd₂) on the empirical formula in Chon's method (Chon et al 2009 IEEE Eng. Med. Biol. Mag. 28 18–23). The results showed that the empirical formula proposed by Chon et al is a good method for analyzing the random signal, but not an appropriate tool for analyzing nonlinear signals, such as the logistic or HRV signals.

Early diagnosis of cerebrovascular disease requires the accurate identification of brain regions with compromised cerebral perfusion pressure (CPP). Current clinical measures of CPP are invasive and lack regional information. Dynamic contrast-enhanced imaging provides a means of looking at regional cerebral hemodynamics. The purpose of this study was to determine if any of the parameters associated with dynamic contrast-enhanced imaging could be used as an index for CPP under graded systemic hypotension in a rabbit model. Cerebral blood flow (CBF), cerebral blood volume, mean transit time (MTT), and cerebrovascular reserve (CVR) were measured using Computed Tomography Perfusion in three groups: normotensive (n = 14), mild hypotensive (n = 9), and moderate hypotensive (n = 6). MTT demonstrated the strongest correlation with CPP (ρ = −0.642, P < 0.05). CBF was the only other parameter to demonstrate a statistically significant correlation (ρ = 0.575, P < 0.05). CVR is gaining momentum for diagnosing cerebrovascular disease; however, the technique requires patients to be given a hemodynamic challenge, which could aggravate symptoms and even trigger stroke. The results of this study suggest that the use of MTT, not requiring hemodynamic manipulation, is more sensitive to subtle changes in CPP, as would occur in the early stages of cerebrovascular disease.

A new arterial distensibility measurement technique was assessed in 100 healthy normotensive subjects. Arterial transmural pressures on the whole right arm were reduced with a 50 cm long cuff inflated to 10, 20, 30 and 40 mmHg. The electrocardiogram, and finger and ear photoplethysmograms were recorded simultaneously. Arm pulse propagation time, pulse wave velocity (PWV) and arterial volume distensibility were determined. With a 40 mmHg reduction in transmural pressure, arm pulse propagation time increased from 61 to 83 ms, PWV decreased from 12 to 8 m s⁻¹ and arterial distensibility increased from 0.102% to 0.232% per mmHg (all P < 0.0001). At all cuff pressures, arterial distensibility was significantly related to resting mean arterial pressure (MAP), diastolic blood pressure (DBP) and age, and for systolic blood pressure at 30 and 40 mmHg (all P < 0.05). At 40 mmHg cuff pressure, arterial distensibility fell by 54% for a MAP increase from 75 to 105 mmHg, 57% for a DBP increase from 60 to 90 mmHg and 47% for an age increase from 20 to 70 years. These changes were more than double than those without cuff pressure. Our technique showed that systemic volume distensibility of the peripheral arm artery reduced with age, with a greater effect at higher external and lower transmural pressures.

Reliable continuous core temperature measurement is of major importance for monitoring patients. The zero heat flux method (ZHF) can potentially fulfil the requirements of non-invasiveness, reliability and short delay time that current measurement methods lack. The purpose of this study was to determine the performance of a new ZHF device on the forehead regarding these issues. Seven healthy subjects performed a protocol of 10 min rest, 30 min submaximal exercise (average temperature increase about 1.5 °C) and 10 min passive recovery in ambient conditions of 35 °C and 50% relative humidity. ZHF temperature (T_zhf) was compared to oesophageal (T_es) and rectal (T_re) temperature. ΔT_zhf–T_es had an average bias ± standard deviation of 0.17 ± 0.19 °C in rest, −0.05 ± 0.18 °C during exercise and −0.01 ± 0.20 °C during recovery, the latter two being not significant. The 95% limits of agreement ranged from −0.40 to 0.40 °C and T_zhf had hardly any delay compared to T_es. T_re showed a substantial delay and deviation from T_es when core temperature changed rapidly. Results indicate that the studied ZHF sensor tracks T_es very well in hot and stable ambient conditions and may be a promising alternative for reliable non-invasive continuous core temperature measurement in hospital.

Real-time monitoring of vital physiological signals is of significant clinical relevance. Disruptions in the signals are frequently encountered and make it difficult for precise diagnosis. Thus, the ability to accurately predict/recover the lost signals could greatly impact medical research and application. We have developed new techniques of signal reconstructions based on iterative retraining and accumulated averaging of neural networks. The effectiveness and robustness of these techniques are demonstrated using data records from the Computing in Cardiology/PhysioNet Challenge 2010. The average correlation coefficient between prediction and target for 100 records of various target signals is about 0.9. We have also explored influences of a few important parameters on the accuracy of reconstructions. The developed techniques may be used to detect changes in patient state and to recognize intervals of signal corruption.

Magnetic induction tomography (MIT) has been proposed for the detection of cerebral oedema and haemorrhagic stroke. Achieving the required phase measurement precision for these applications is however a major technical challenge. A critical component within an MIT system is the detector amplifier and for this role an ultra-phase-stable, low noise instrumentation amplifier has been developed. The design of the amplifier is described and (i) the results of simulations and measurements of the amplifiers phase stability versus temperature and (ii) measurements of the phase noise and drift performance of the amplifier within a single-channel magnetic induction spectroscopy system are provided and discussed. For a 10 MHz signal the amplifier, with a gain of 21, displayed an average change in the measured phase of its output of just −0.1 ± 0.6 m° °C^–1 as the ambient temperature was varied between 35 and 50 °C, demonstrating a level of phase stability approaching that required for potential biomedical applications such as the detection of cerebral haemorrhage.

This paper presents a new family of indices for the frequency domain analysis of heart rate variability time series that do not need any frequency band definition. After proper detrending of the time series, a cumulated power spectrum is obtained and frequencies that contain a certain percentage of the power below them are identified, so median frequency, bandwidth and a measure of the power spectrum asymmetry are proposed to complement or improve the classical spectral indices as the ratio of the powers of LF and HF bands (LF/HF). In normal conditions the median frequency provides similar information as the classical indices, while the bandwidth and asymmetry can be complementary measures of the physiological state of the tested subject. The proposed indices seem to be a good choice for tracking changes in the power spectrum in exercise stress, and they can guide in the determination of frequency band limits in other animal species.

Hypertensive pregnancy disorders affect 6% to 8% of all pregnancies and can result in severe complications for the mother and the foetus of which pre-eclampsia (PE) has the worst perinatal outcome. Several studies suggested that the autonomic nervous system plays an important role in the process of developing hypertensive pregnancy disorders, especially PE. The aim of this retrospective study was to investigate whether women with PE could be differentiated from women with various other hypertensive pregnancy disorders, by employing an enhanced Poincaré plot analysis (PPA), the segmented Poincaré plot analysis (SPPA), to their beat-to-beat interval and blood pressure signals. Sixty-nine pregnant women with hypertensive disorders (29 PE, 40 with chronic or gestational hypertension) were included. The SPPA as well as the traditional PPA found significant differences between PE and other hypertensive disorders of diastolic blood pressure (p < 0.001 versus p < 0.001) but only the SPPA method revealed significant differences (p < 0.001) also of the systolic blood pressure. Further on, linear discrimination analysis demonstrated that indices derived from SPPA are more suitable for differentiation between chronic and gestational hypertension and PE than those from traditional PPA (area under the ROC curve 0.85 versus 0.69). Therefore this procedure could contribute to the differential diagnosis of hypertensive pregnancy disorders.

Central blood pressure (CBP) has been established as a relevant indicator of cardiovascular disease. Despite its significance, CBP remains particularly challenging to measure in standard clinical practice. The objective of this study is to introduce pulse wave-based ultrasound manometry (PWUM) as a simple-to-use, non-invasive ultrasound-based method for quantitative measurement of the central pulse pressure. Arterial wall displacements are estimated using radiofrequency ultrasound signals acquired at high frame rates and the pulse pressure waveform is estimated using both the distension waveform and the local pulse wave velocity. The method was tested on the abdominal aorta of 11 healthy subjects (age 35.7±16 y.o.). PWUM pulse pressure measurements were compared to those obtained by radial applanation tonometry using a commercial system. The average intra-subject variability of the pulse pressure amplitude was found to be equal to 4.2 mmHg, demonstrating good reproducibility of the method. Excellent correlation was found between the waveforms obtained by PWUM and those obtained by tonometry in all subjects (0.94 < r < 0.98). A significant bias of 4.7 mmHg was found between PWUM and tonometry. PWUM is a highly translational method that can be easily integrated in clinical ultrasound imaging systems. It provides an estimate of the pulse pressure waveform at the imaged location, and may offer therefore the possibility to estimate the pulse pressure at different arterial sites. Future developments include the validation of the method against invasive estimates on patients, as well as its application to other large arteries.

The autonomic regulation is non-invasively estimated from heart rate variability (HRV). Many methods utilized to assess autonomic regulation require stationarity of HRV recordings. However, non-stationarities are frequently present even during well-controlled experiments, thus potentially biasing HRV indices. The aim of our study is to quantify the potential bias of spectral, symbolic and entropy HRV indices due to non-stationarities. We analyzed HRV series recorded in healthy subjects during uncontrolled daily life activities typical of 24 h Holter recordings and during predetermined levels of robotic-assisted treadmill-based physical exercise. A stationarity test checking the stability of the mean and variance over short HRV series (about 300 cardiac beats) was utilized to distinguish stationary periods from non-stationary ones. Spectral, symbolic and entropy indices evaluated solely over stationary periods were contrasted with those derived from all the HRV segments. When indices were calculated solely over stationary series, we found that (i) during both uncontrolled daily life activities and controlled physical exercise, the entropy-based complexity indices were significantly larger; (ii) during uncontrolled daily life activities, the spectral and symbolic indices linked to sympathetic modulation were significantly smaller and those associated with vagal modulation were significantly larger; (iii) while during uncontrolled daily life activities, the variance of spectral, symbolic and entropy rate indices was significantly larger, during controlled physical exercise, it was smaller. The study suggests that non-stationarities increase the likelihood to overestimate the contribution of sympathetic control and affect the power of statistical tests utilized to discriminate conditions and/or groups.

We have applied principles of statistical signal processing and nonlinear dynamics to analyze heart rate time series from premature newborn infants in order to assist in the early diagnosis of sepsis, a common and potentially deadly bacterial infection of the bloodstream. We began with the observation of reduced variability and transient decelerations in heart rate interval time series for hours up to days prior to clinical signs of illness. We find that measurements of standard deviation, sample asymmetry and sample entropy are highly related to imminent clinical illness. We developed multivariable statistical predictive models, and an interface to display the real-time results to clinicians. Using this approach, we have observed numerous cases in which incipient neonatal sepsis was diagnosed and treated without any clinical illness at all. This review focuses on the mathematical and statistical time series approaches used to detect these abnormal heart rate characteristics and present predictive monitoring information to the clinician.