310 likes | 475 Views
VLSI and Embedded Systems Conference, 5-9 Jan 2014, IIT Bombay, India ( VLSIDES14 ) Session: B-2 Embedded Platform, Venue: VMCC -21, Session Time: 4:30 pm to 6:30 pm An Embedded System Design for a Synchronous Demodulation Based Noninvasive Bioimpedance Sensor P. C . Pandey. IIT Bombay
E N D
VLSI and Embedded Systems Conference, 5-9 Jan 2014, IIT Bombay, India (VLSIDES14) Session: B-2 Embedded Platform, Venue: VMCC-21, Session Time: 4:30 pm to 6:30 pm An Embedded System Design for a Synchronous Demodulation Based Noninvasive Bioimpedance Sensor P. C. Pandey IIT Bombay 07/Jan/2014 <pcpandey@ee.iitb.ac.in> http://www.ee.iitb.ac.in/~pcpandey
VLSI and Embedded Systems Conference, 5-9 Jan 2014, IIT Bombay, India (VLSIDES14) Session: B-2 Embedded Platform, Venue: VMCC-21, Session Time: 4:30 pm to 6:30 pm P. C. Pandey: An embedded system design for a synchronous demodulation based noninvasive bioimpedance sensor (invited talk) Abstract: A long-duration noninvasive monitoring of bioimpedance has the potential of serving as a low-cost diagnostic tool and monitoring device in several medical applications, e.g. impedance cardiography (sensing of variation in thoracic impedance to estimate cardiac output and some other hemodynamic parameters), pneumography (sensing of respiratory parameters), plethysmography for sensing peripheral blood circulation, glottography (for sensing movement of vocal chords), etc. These instruments pass an alternating current of high frequency and low amplitude through a pair of appropriately placed pair of electrodes, an amplifier to sense the resulting amplitude modulated voltage across the same or another pair of appropriately placed electrodes, a demodulator to detect the impedance signal, and signal processing for obtaining the desired parameters. An embedded system design approach is used to develop a body-worn device to be used for monitoring the clinically important physiological parameters during critical care, for ambulatory recording for early diagnosis of cardiovascular disorders and for post-operative care, for monitoring of physiological parameters for use in sports medicine, and as a low-cost diagnostic aid. It senses the basal value and time-varying component of the impedance waveform, with settable excitation frequency and with very low noise and demodulation related distortions. A microcontroller and an impedance converter chip are used for stable sinusoidal source with programmable frequency control and a digital synchronous demodulation. A voltage-to-current converter with balanced outputs is designed using two operational trans-conductance amplifiers for current excitation. The sensed voltage is added with a sinusoidal voltage obtained from the excitation source and with digitally controlled amplitude and polarity to increase its modulation index before digital synchronous demodulation and for baseline correction of the sensed impedance signal. Two digital potentiometers have been used to provide independent control over current excitation and baseline correction. Synchronous digital demodulation in the impedance converter chip gives real and imaginary part of the impedance. An isolated RS232 interface is provided to set the parameters and to acquire the sensed impedance signal. Dr. P. C. Pandey, Professor, Electrical Engineering, IIT Bombay EE Dept, IIT Bombay, Powai Mumbai 400076, India <pcpandey@ee.iitb.ac.in> http://www.ee.iitb.ac.in/~pcpandey
Outline Introduction Design Approach Hardware & Software Test & Results Summary Reference HitendraSahu: “Sensing of impedance cardiogram using synchronous demodulation”, M. Tech. dissertation, Biomedical Engineering, Indian Institute of Technology Bombay, June 2013.
Noninvasive Monitoring of Bioimpedance • Low-cost diagnostic tool • Monitoring device • Some Applications • Impedance cardiography: sensing of variation in thoracic impedance to estimate cardiac output & some other hemodynamic parameters • Pneumography: sensing of respiratory parameters • Plethysmography: sensing of peripheral blood circulation • Glottography: sensing movement of vocal chords during speech production
Instrumentation for Bioimpedance Sensing • Passing an alternating current of high frequency and low amplitude through a pair of appropriately placed pair of electrodes • Amplifier to sense the resulting amplitude modulated voltage across the same or another pair of appropriately placed electrodes • Demodulator to detect the impedance signal • Signal processing for obtaining the desired parameters
Example: Impedance Cardiograph • ICG blocks • AC excitation current • Voltage sensing amp. • Demodulator • Baseline correction • ECGextractor • Operation • Excitation current: 20 - 100 kHz, < 5 mA • Amplitude demodulation of the sensed voltage: Z(t) with basal impedance (20 − 200 Ω)& time-varying component (< 0.2 Ω) • ICG: − dZ/dt, processed with ECG as the reference.
Objective • To develop a body-worn bioimpedance sensing device for • Monitoring the clinically important physiological parameters during critical care (multi-channel signal acquistion & processing) • Ambulatory recording for early diagnosis of cardiovascular disorders and for post-operative care (recording in the presence of motion artifacts) • Monitoring of physiological parameters for use in sports medicine (recording in the presence of external interference, strong respiratory and motion artifacts) • Low-cost diagnosis (low distortion & high sensitivity)
Design Approach • Digital synchronous demodulation for noise and interference rejection • Circuit for increasing the modulation index of the waveform to increase the sensitivity and dynamic range Basic Blocks • Microcontroller “Microchip PIC24FJ64GB04” • Impedance converter chip “Analog Devices AD5933” • V-to-I convertor and amplitude control • Voltage sensing amplifier and baseline correction • PC-based GUI with isolated serial communication for setting parameters and data acquisition
Impedance converter AD5933 Features • Excitation voltage generator & digital synch. demodulator • Programmable voltage with a settable frequency up to 100 kHz • Impedance measurement range from 1 kΩto 10 MΩ • Internal system clock • DC rejection, error averaging, phase measurement • Accuracy: ± 0.5%. • I2C interface with a data rate of 100 kHz Adaptations needed for bioimpedance sensing • Measurement using current excitation • Time-varying measurement • Dynamic range extension and sensitivity selection
Design using the impedance converter chip • with on-chip sinusoidal source & DFT for synchronous digital demodulation
Digital pot. AD8400 (U3, U7) used for controlling the amplitudes the excitation current and baseline correction voltage. • Total resistance 1 K with 8 bit resolution. • Wiper position changed via SPI interface. • Supply range : 2.6 – 5.5 V.
V-to-I converter • Voltage generated from AD5933 is converted to a current • Operational trans-conductance amplifier (OTA) based V – I converter used • “TI OPA861” from used as OTA which offers : • gm is 90 mA/V at a quiescent current of 4.7 mA • +5 V single supply operation • Bandwidth 80 MHz
V – I converter with balanced current outputs • Complimentary balanced current output generated by using two OTA chips • Quiescent current set byR1 and R3 connected at base terminals • Output current can is controlled by changing R2 or quiescent current
Voltage sensing amplifier • Instr. amp. INA155 for amplifying the sensed voltage • BW: 5.5 MHz • Gain: 10 – 50 • Slew rate 6.5 V/µs • Supply: 2.6 – 5.5 V • High pass filter cut-off : 16 kHz
where , where K= ),β, and α are potentiometer ratios from U3 and U7 respectively • Baseline correction • Subtracting a sinusoidal reference voltage from the sensed voltage • Amplitude and polarity of the correction voltage digitally controlled by varying digital pot (U7) ratio between 0.25 to 0.75 • Baseline correction output tracked by microcontroller using ADC. • Potentiometer ratio is controlled digitally via SPI interface
Demodulation • Samples the sensed signal at 1.04 MHz. • 1024-point DFT is calculated at the excitation frequency. • The DFT algorithm returns a real (R) and imaginary (I) data values. • Data values are transferred to microcontroller over an I2C. • Impedance values send to PC over UART or USB.
Microcontroller • 44-pin PIC24F64GB004 used • Supply range : 3.0 – 3.6 V • 16 MHz clock • 64 KB program memory, • 8 KB RAM • Single channel 10 bit ADCs • UART module • USB module • SPI module • I2C module
Power supply features • Separate analog & digital supplies of 3.3 V & 5 V. • Analog reference of 1.6 V generated by MCP6021. • LDO MCP1802 used as voltage regulator IC. • Input to the LDO from a DC-DC converter LM2622. • Input to the DC-DC converter: 3.6-5.5 V. • Li-ion charge control IC MCP73833 used for battery charging. • Total current consumption ~60 mA. • Low battery indication. • Provision for powering through USB.
Signal acquisition interface LabWindowsCVI software for signal acquisition using RS232
Test & Results • Exc.: 65.5 kHz, 0.9 mA • Lin. range: up to 400 Ω A) Voltage sensing amplifier: output linearity B) Interference Significant only over a b.w. of 3 kHz
C) Automatic Sensitivity Adjustment Voltage sensing amplifier output vs test resistances for excitation current of 0.6 − 1.5 mA, set by varying β
D) Validation using thoracic impedance simulator Excitation: 0.6 mA, 65.56 kHz Simulator settings: R = 49 Ω, ∆R = 0.5Ω , f = 1 Hz Sampling freq.: 200 Hz Excitation: 0.6 mA, 65.56 kHz Simulator settings: R = 20 Ω, ∆R = 0.8 Ω, f = 0.1 Hz Sampling freq.: 10 Hz
Excitation: 0.6 mA, 65.56 kHz Simulator settings: R = 30 Ω, ∆R = 0.8 Ω, f = 0.1 Hz Sampling freq. : 200 Hz Excitation: 0.6 mA, 65.56 kHz Simulator settings : R = 19 Ω, ∆R = 0.5 Ω , f = 5 Hz Sampling freq.: 200 Hz
Summary Developed A bioimpedance sensor using an impedance converter chip using digital synchronous demodulation Further work • Median filtering for further carrier ripple rejection without smearing transitions • Adaptation for for specific applications • Integration with the signal processing software • Field testing
References • [1] R. P. Patterson, "Fundamentals of impedance cardiography," IEEE Eng. Med. Biol. Mag., vol. 8, no. 1, pp. 35-38, 1989. • [2] L. E. Baker, "Applications of impedance technique to the respiratory system," IEEE Eng. Med. Biol. Mag., vol. 8, no. 1, pp. 50–52, 1989. • [3] L. E. Baker, "Principles of impedance technique," IEEE Eng. Med. Biol. Mag., vol. 8, no. 1, pp. 11–15, 1989. • [4] H. H. Woltjer, H. J. Bogaard, and P. M. J. M. de Vries, “The technique of impedance cardiography,” Euro. Heart J., vol. 18, no. 9, pp. 1396–1403, 1997. • [5] M. D. Desai, “Development of an impedance cardiograph,” M. Tech. dissertation, Biomedical Engineering,, IIT Bombay, 2012. • [6] H. Sahu: “Sensing of impedance cardiogram using synchronous demodulation”, M. Tech. dissertation, Biomedical Engineering, IIT Bombay, June 2013.