Table of contents

This study assessed the clinical and practical value of angioplasty balloon pressure–volume data, obtained by a computer-controlled balloon inflation device, during standard percutaneous coronary intervention (PCI) procedures. Stent deployment was studied for 57 lesions in 34 patients. Angiographic predilation data were available in 9 patients. Vessel lumen improvement with pre-dilation was determined using the area difference under the pressure–volume curves of two consecutive inflations and compared to the angiographically determined lumen improvement. Stent opening pressure, the inflation volume needed to unfold the stent and the shape of the pressure–volume curve were assessed for differences between stent sizes (diameter, length) and manufacturer types. A Pearson correlation of 0.8 between the pressure–volume area difference and the angiographic lumen improvement was significant (p = 0.01), confirming that pressure–volume curves are useful in assessing lumen improvement after pre-dilation. There was a significant difference in stent opening pressure between stent types but not between different stent sizes (length, diameter). The inflation volume measured during the unfolding process of the stent correlated with the calculated stent lumen when deployed (Pearson correlation: 0.65, p = 0.001). The shape of the pressure–volume curve during stent deployment illustrated differences between different stent sizes (length, diameter) during inflation. Pressure–volume curves obtained from an automatic balloon inflation device have shown their usefulness in providing additional feedback about lumen improvement and the mechanical characteristics and quality of stent deployment.

A time-varying parametric spectrum estimation method for analysing non-stationary heart rate variability signals is presented. As a case study, the dynamics of heart rate variability during an orthostatic test is examined. In this method, the non-stationary signal is first modelled with a time-varying autoregressive model and the model parameters are estimated recursively with a Kalman smoother algorithm. The benefit of using the Kalman smoother is that the lag error present in a Kalman filter, as well as in all other adaptive filters, can be avoided. The spectrum estimates for each time instant are then obtained from the estimated model parameters. Statistics of the obtained spectrum estimates are derived using the error propagation principle. The obtained spectrum estimates can further be decomposed into separate components and, thus, the time variation of low- and high-frequency components of heart rate variability can be examined separately. By using the presented method, high resolution time-varying spectrum estimates with no lag error can be produced. Other benefits of the method are the straightforward procedure for evaluating the statistics of the spectrum estimates and the option of spectral decomposition.

Alzheimer's disease (AD) is the most common neurodegenerative disorder. Although a definite diagnosis is only possible by necropsy, a differential diagnosis with other types of dementia and with major depression should be attempted. The aim of this study was to analyse the electroencephalogram (EEG) background activity of AD patients to test the hypothesis that the regularity of the AD patients' EEG is higher than that of age-matched controls. We recorded the EEG from 19 scalp electrodes in 11 AD patients and 11 age-matched controls. Two different methods were used to estimate the regularity of the EEG background activity: spectral entropy (SpecEn) and sample entropy (SampEn). We did not find significant differences between AD patients and control subjects' EEGs with SpecEn. On the other hand, AD patients had significantly lower SampEn values than control subjects (p < 0.01) at electrodes P3, P4, O1 and O2. Our results show an increase of EEG regularity in AD patients. These findings suggest that nonlinear analysis of the EEG with SampEn could yield essential information and may contribute to increasing the insight into brain dysfunction in AD in ways which are not possible with more classical and conventional statistical methods.

Near infrared spectroscopy (NIRS) is a non-invasive method to monitor cerebral haemodynamics. Used either alone or in combination with other non-invasive methods such as transcranial Doppler sonography, this technique is well suited for use in cerebrovascular research in ageing. Reproducibility of NIRS, however, has only been determined in neonates and adults. We applied controlled desaturation (the O₂-method) to measure the cerebral blood volume (CBV) with NIRS in 16 healthy subjects aged 65 to 88. This method uses deoxygenated haemoglobin (the concentration of which is manipulated by desaturation) as an intravascular tracer for NIRS. We determined repeatability (between tests interval: 2 min), short-term reproducibility (intervals of 20 and 40 min) and long-term reproducibility (interval > 2 weeks). We found a coefficient of variation (CV) of 12.5% for repeatability and a CV of 11.7% for short-term reproducibility. The CV for long-term reproducibility was 15%. We conclude that NIRS can reproducibly measure CBV in subjects aged 65 and older, using the O₂-method. In this group of healthy subjects, this method was well tolerated.

Impedance spectroscopy has been proposed as a method of monitoring mucosal injury due to hypoperfusion and ischemia in critically ill patients. The present study characterizes human gastric impedance spectral changes under gastric hypoperfusion in patients undergoing cardiovascular surgery, and evaluates spectral differences between patients with no evidence of gastric ischemia and complications, and patients who developed ischemia and complications. Cole and Kun parameters were calculated over time to characterize changes as tissue injury progresses. Gastric ischemia was determined by air tonometry. Impedance spectroscopy spectra were obtained from 63 cardiovascular surgery patients. The recorded spectra were classified into three groups: group 1 for patients without ischemia or complications, group 2 for patients with a short period of ischemia (less than 2 h) and group 3 for patients with more than 4 h of gastric ischemia and complications. Two distinct dispersion regions of the spectra centered at about 316 Hz and 215 kHz become clearer as tissue damage develops. The average spectrum in group 3 shows a significant difference in tissue impedance at all frequencies relative to group 1. The parameters obtained for human gastric tissue show significant changes that occur at different times and at different frequencies as ischemia progresses, and could be correlated with patient outcome. This confirms our hypothesis that hypoperfusion and ischemia cause evident changes in the impedance spectra of the gastric wall. Therefore, this technology may be a useful prognostic and diagnostic monitoring tool.

The ¹³C-octanoate breath test has not achieved universal acceptance for the measurement of solid phase gastric emptying, largely because the results are not comparable with those from direct methods such as scintigraphy. To convert breath-test data to their scintigraphic equivalent scale corrections are applied which have been obtained from population studies, but there is no guarantee that these are applicable in all cases. We propose an alternative method applicable on an individual basis based upon a simple physiological model which does not require any change to the breath-test protocol. It is demonstrated by comparison with scintigraphy and the octanoate saliva test. Results from an existing dataset of simultaneous breath test, saliva test and scintigraphic determinations of solid phase gastric emptying from nine healthy subjects were re-analysed. The corrected breath tests gave gastric emptying curves which were shown to be not significantly different to those obtained from scintigraphy. The method provides a simple but effective way of generating gastric emptying curves from breath-test data that are directly comparable with direct measurement methods, which is advantageous since it allows the whole of the gastric emptying profile to be generated, not just values for the lag phase and half-emptying times.

In electrical impedance tomography (EIT) the electrode structure and parameters significantly influence measurement sensitivity and image quality, so how to optimize the electrode structure and parameters is one of the key problems in research today. This paper presents a method to optimize the EIT electrode structure and parameters based on coercive equipotential node models. The coercive equipotential mode of the compound electrode has been established based on that of the line electrode. A simulation study for the line electrode and the compound electrode of EIT has been made on a simulation software platform. The influences of different electrode structures and parameters on measurement sensitivity and the image reconstruction quality are studied. For line electrode simulation studies, two important conclusions are drawn. First, a narrower electrode is helpful in improving the imaging quality. Second, although it is known that a wider electrode is beneficial in decreasing the contact impedance, using a too wide electrode causes the measurement sensitivity to decrease. Furthermore the electrode width that leads to the best measurement sensitivity is different for different measurement depths. The compound electrode has four parameters: the excitation electrode width, the measurement electrode width, the space between the excitation electrode and the measurement electrode, and the distance between two adjacent compound electrodes. These parameters have mutual restrictions and complex influences on each other. It is unwise to optimize the design of a compound electrode by only using the overlay rate of electrodes. A simulation study of EIT electrode structure and parameter influences can be carried out according to this paper to determine the optimum design of the electrode structure and its parameters.

The effect of foetal heart rate (FHR) acquisition mode on linear and nonlinear parameters is still largely unknown. In 33 normal labouring women, FHR signals were acquired simultaneously by an external ultrasound sensor applied to the maternal abdomen and an internal scalp electrode, in the minutes preceding delivery. For each case, the initial and final 5, 10 and 20 min segments were analysed, considering FHR signals at a frequency of 4 Hz (the frequency at which they are transmitted by the majority of commercialized foetal monitors). Several time and frequency domain linear and nonlinear FHR indices were computed in these segments, namely mean FHR, very low frequency (VLF), low frequency (LF), high frequency (HF), approximate entropy (ApEn) and sample entropy (SampEn). Parametric confidence intervals, statistical tests and correlation coefficients were calculated in order to evaluate the effect of internal versus external FHR monitoring modes on the considered indices. The whole evaluation was repeated using FHR signals at a frequency of 2 Hz. Most time domain linear indices were similar with external and internal monitoring in the initial and final segments of the tracings. However, linear frequency domain indices were poorly correlated in the final segments and had significantly different mean values in the initial segments. Nonlinear indices were significantly different in both initial and final segments. The correlation between 4 and 2 Hz sampled parameters was high for both linear and nonlinear indices (most correlation coefficient values ranging between 0.95 and 1) but nonlinear index values were significantly higher at 2 Hz. In conclusion, the mode used to acquire FHR signals and the sampling rate employed can significantly affect most FHR indices.