Blagoy P. Iliev

On-line blood viscosity monitoring in vivo with a central venous catheter, using electrical impedance technique.

Blood viscosity is an important determinant of microvascular hemodynamics and also reflects systemic inflammation. Viscosity of blood strongly depends on the shear rate and can be characterized by a two parameter power-law model. Other major determinants of blood viscosity are hematocrit, level of inflammatory proteins and temperature. In-vitro studies have shown that these major parameters are related to the electrical impedance of blood. A special central venous catheter was developed to measure electrical impedance of blood in-vivo in the right atrium. Considering that blood viscosity plays an important role in cerebral blood flow, we investigated the feasibility to monitor blood viscosity by electrical bioimpedance in 10 patients during the first 3 days after successful resuscitation from a cardiac arrest. The blood viscosity-shear rate relationship was obtained from arterial blood samples analyzed using a standard viscosity meter. Non-linear regression analysis resulted in the following equation to estimate in-vivo blood viscosity (Viscosity(imp)) from plasma resistance (R(p)), intracellular resistance (R(i)) and blood temperature (T) as obtained from right atrium impedance measurements: Viscosity(imp)=(-15.574+15.576R(p)T)SR ((-.138RpT-.290Ri)). This model explains 89.2% (R(2)=.892) of the blood viscosity-shear rate relationship. The explained variance was similar for the non-linear regression model estimating blood viscosity from its major determinants hematocrit and the level of fibrinogen and C-reactive protein (R(2)=.884). Bland-Altman analysis showed a bias between the in-vitro viscosity measurement and the in-vivo impedance model of .04 mPa s at a shear rate of 5.5s(-1) with limits of agreement between -1.69 mPa s and 1.78 mPa s. In conclusion, this study demonstrates the proof of principle to monitor blood viscosity continuously in the human right atrium by a dedicated central venous catheter equipped with an impedance measuring device. No safety problems occurred and there was good agreement with in-vitro measurements of blood viscosity.

An interface circuit for R-C impedance sensors with a relaxation oscillator

A simple interface circuit for impedance sensors based on a relaxation oscillator is presented. The measurement strategy and the principle of operation are discussed. The circuit is intended for measuring impedance, which can be represented as a capacitor and a resistor in series. By means of a four-signal measurement technique, continuous self-calibration is achieved. The output signals for both components of the unknown impedance are time periods, and for calculating their values only one reference capacitor is needed. A measurement set-up is presented with which the new method was tested. The experimental results prove that with this new technique, very high resolution for both components of the measured impedance can be achieved-better than 0.1 /spl Omega/ for R. and better than 0.1 pF for C/sub x/.

Extending the Limits of a Capacitive Soil-Water-Content Measurement

The capacitive component of soil impedance is a good measure for the water content. For long time people use rod-shaped pairs of electrodes to measure electrically the water content in natural- and artificial soil. Especially in high-conductive mediums, with conductivity as high as 15 mS/cm, this technique is not able to acquire the capacitive component of the impedance reliably. Because of the necessary usage of high measurement frequencies, physical effects, such as skin effect limit the applicability of long rod-shaped electrode pairs. Having analyzed this problem a new electrodes structure has been built and tested to reduce this unwanted physical effect thus enabling reliably water-content measurements

A novel technique to measure two independent components of impedance sensors with a simple relaxation oscillator

A simple interface circuit for impedance measurement based on a relaxation oscillator is presented. The measurement strategy is revealed and the principle of operation is discussed. It is intended for measuring impedance, which can be presented with a capacitor and resistor in series. By means of a four-signal measurement technique a continuous self-calibration is achieved. The output signals for both components of the unknown impedance are time-periods and for calculating their values only one reference capacitor is needed. A measurement setup is presented with which the new method was tested. The experimental results prove that with this new technique very high sensitivity for both components of the measured impedance can be achieved.

Integrated sensor arrays for bioluminescence and fluorescence bio-chemical analysis

We present on-chip luminescence and fluorescence bio-chemical analysis, using integrated photodiodes. The detectors and the read-out electronics are implemented on a silicon substrate using standard CMOS processing. The photosensitive structures result from two-stacked PN junctions and an (optional) optical filter. The bioluminescent analyses are based on a light producing reaction - the conversion of ATP (adenosine triphosphate) molecules to AMP - catalyzed by the enzyme luciferase. The obtained results for three different initial concentrations of ATP molecules, in ATP consuming reactions, are presented. Initial fluorescent measurements have been conducted, based on the enzyme protein tyrosine phosphatase (PTP1B) using molecular probes DiFMUP (UV excitable). An enzyme solution (500 pg//spl mu/l) was mixed with DiFMUP. The reaction product DiFMU exhibits excitation/emission maxima of /spl sim/358/455 nm. The undesired excitation (UV) light was filtered out with. an integrated on-chip high pass filter with wavelength cut-off at 400 nm.

In-vivo Blood Characterization System

A blood-impedance measurement system is presented. It is intended for in-vivo analysis of hematocrit and blood viscosity. Using a specially designed central-venous catheter, blood impedance and blood temperature is on line examined. The measurements are performed inside the right atrium, where a strong intracavitary ECG signal is recorded as well. The impedance is measured with sinusoidal signals at five discrete frequencies in the range from 20 kHz to 1.2 MHz. The temperature is measured with a tiny thermistor incorporated in the catheters tip. From the various blood-impedance components, the measurement system derives plasma resistance, cell membrane capacitance and cell interior resistance. Proper catheter positioning and processing of the measured impedance components are guided by the ECG signal, sensed with the same impedance sensor. A wide variety of biocompatibility issues has thoroughly been identified and related problems have been solved

Sensorized nanoliter reactor chamber for DNA multiplication

This paper presents thermal analysis verification of a sensorized 50 nl reactor chamber for DNA amplification based on PCR (polymerase chain reaction). The reactor is equipped with an integrated heater, temperature sensor and a photo detector for real time detection. Through micromachining, the thermal capacity of each chamber is minimized, enabling rapid PCR cycling. The proposed structure was implemented on a silicon substrate using a standard CMOS process and postprocessing. The chambers have a bottom area of 500/spl times/500 /spl mu/m/sup 2/ and a pitch of 1 mm. An array of 96 reactors can be formed on a square centimeter. In order to reach the required PCR temperature levels (55/spl deg/C, 75/spl deg/C and 92/spl deg/C) dedicated electronics, based on a proportional-integral (PI) controller were designed and built. The system is capable of stabilizing the temperature of the reactor and performing a temperature sweep up to 100/spl deg/C.

A multi-period interface system for impedance measurements

A novel multi-period interface system for impedance measurement, based on a first-order oscillator, is presented. The system is designed for sterility testing of aseptically packed food products by noninvasive measurement of the conductivity of the packaged food. In this application the measured impedance can be modeled as a resistor, representing the conductivity of the food in series with a capacitor, representing the electrical behavior of the walls of the food container. The measurement range is: 50 pF to 220 pF for the capacitive component and 10 /spl Omega/ to 150 /spl Omega/ for the resistive component. By applying auto-calibration, a relative error of 0.3% for the resistive component R/sub x/ and 0.1% error for the capacitive component C/sub x/ is achieved. The output signal is a period-modulated square wave signal, which can be processed directly by a microcontroller, without any additional interface circuit. To calculate the values for R/sub x/ and C/sub x/, a data processing algorithm has been developed. The details of a prototype design have been discussed together with the experimental results.

A Sensor Interface System for Measuring the Impedance (C x , R x ) of Soil at a Signal Frequency of 20MHz

A 20 MHz impedance-measurement system for measuring the equivalent parallel capacitance Cx and the shunting resistance Rx in artificial soil is presented. For water-content measurements in artificial soil, with a special two-electrode probe, the Cx values vary from 1 pF up to 30 pF. These small capacitances have to be measured in the presence of a shunting resistor Rx, with values ranging from 22 Omega up to 1 kOmega. The measurements have been performed with a sinusoidal signal with a frequency of 20 MHz. Based on an existing principle an improved interface has been developed, which enables accurate measurement of both the capacitive component Cx and its shunting resistance Rx. The high performance of the improved system has been obtained by implementing two direct digital synthesizers (DDS), which are used to reduce the nonlinearity in both the amplitude-ratio measurement and the phase measurement. Also so-called open/short compensation has been applied for compensating the parasitics in the measurement setup. Experimental results show that for Cx with a full-scale range of 30 pF the relative error amounts to 1% and 1.5% for shunting resistors of 820 Omega and 47 Omega, respectively. For the Rx component, for a full-scale range of 1 kOmega, the error is less than 1%.