DESIGN OF SENSORY SYSTEM FOR BREATH ANALYSISDESIGN OF SENSORY SYSTEM FOR BREATH ANALYSISDESIGN OF SENSORY SYSTEM FOR BREATH ANALYSISDESIGN OF SENSORY SYSTEM FOR BREATH ANALYSIS1 1 2 3Yingkan Lin1 Milutin Stanaćević1 Sanford R Simon2 and Pelagia Irene Gouma3Yingkan Lin1, Milutin Stanaćević1, Sanford R. Simon2 and Pelagia‐Irene Gouma3g , , g

Departments of Electrical & Computer Engineering1 Biochemistry2 Pathology2 and Material Science & Engineering3Departments of Electrical & Computer Engineering , Biochemistry , Pathology and Material Science & EngineeringStony Brook University Stony Brook NY 11794Stony Brook University, Stony Brook, NY 11794y y, y ,

k d d dk d d d d ld lBackground and IntroductionBackground and Introduction Discrete Readout System ImplementationDiscrete Readout System Implementation Sensor baseline resistance range: 1kΩ ~ 100MΩBackground and IntroductionBackground and Introduction Discrete Readout System ImplementationDiscrete Readout System Implementation Sensor baseline resistance range: 1kΩ  100MΩgg y py p0 05% ~ 10% of sensor resistance change detectable0.05% ~ 10% of sensor resistance change detectableT l d i 166dBH dh ld t bl i l FENO b th l h b d i dThe la k of ine pensi e sensor te hnolo that o ld dete t and monitor the on entration Total dynamic range: 166dBHandheld, portable wireless FENO breath analyzer has been designed.The lack of inexpensive sensor technology that would detect and monitor the concentration  y g, p y g

l d b l bl f db k fp gy

of a single gas with high specificity and at low trace concentrations in the presence of Power Consumption: 400 μWSignal conditioning comprises baseline compensation, variable feedback resistance forof a single gas with high specificity and at low trace concentrations, in the presence of  Power Consumption: 400 μWSignal conditioning comprises baseline compensation, variable feedback resistance for l f h d i d 2nd d l filnumerous interfering compounds has presented a main obstacle in advancement of the

Chip area: 1 4mm x 0 8mmcontrol of the dynamic range and 2nd order low‐pass filter.numerous interfering compounds, has presented a main obstacle in advancement of the Chip area: 1.4mm x 0.8mmy g p

diagnostics using breath analysisFabricated in 0 5μm CMOS technologyProposed signal processing techniques are implemented on a micro‐controller.diagnostics using breath analysis. Fabricated in 0.5μm CMOS technologyProposed signal processing techniques are implemented on a micro controller.

Sensor resistance range: 100Ω ~ 100MΩSensor resistance range: 100Ω  100MΩTask: design of a sensory system for detection and discrimination of nitric oxide in exhaled

Size: 7 5cm x 7 5cmTask: design of a sensory system for detection and discrimination of nitric oxide in exhaled

Size: 7.5cm x 7.5cmbreath and measurement of the trace concentration for affordable and noninvasive medicalPower consumption 168mW (360mW when Bluetooth is on)

breath and measurement of the trace concentration for affordable and noninvasive medicalPower consumption: 168mW (360mW when Bluetooth is on)diagnostics technologydiagnostics technology.

Obj i d l d l FENO i h ldObjective: to develop a stand‐alone FENO measurement micro‐system that would operate asj p y pt b t d li bl l d ia easy to use, robust and reliable personal device.y , p

Layout of the integrated readoutLayout of the integrated readout h lChallenges in the design of the system: system in 0.5μm CMOS technologyChallenges in the design of the system:

l l d ll / f dy μ gy

a simple, low cost and small area gas sensor/sensor array specific to nitric oxidea simple, low cost and small area gas sensor/sensor array specific to nitric oxidehi i th h t t d th l t t bilit f th achieving the short‐term and the long‐term stability of the sensor response

Experimental ResultsExperimental Resultsg g y p

hi h iti it d l d t i it Experimental ResultsExperimental Results high sensitivity and low power readout circuit xperimental Resultsxperimental Resultsg y plow power heating element low power heating element

Electronic package 500ppbHeater

500ppbHeaterSensor substrate 300ppb300ppb

Block diagram of the wireless prototype Discrete 200ppbBlock diagram of the wireless prototype. Discrete  200ppb

implementation of signal conditioningimplementation of signal conditioning 100ppbtop view side view 100ppbtop view side view

Sensor and heater assemblySensor and heater assembly.

Sensory System OverviewSensory System OverviewSensory System OverviewSensory System OverviewSe so y Syste O e eSe so y Syste O e eLimits on sensitivity:Limits on sensitivity:

electrical noiseelectrical noise

T l f th NO i R f th NO i t f large baseline resistance that can be few orders of magnitude higher than the actual Temporal response of the NO sensing  Response of the NO sensing system for large baseline resistance that can be few orders of magnitude higher than the actual  p p gsystem for different concentrations of

gdifferent concentrations of NO for shortsensor response system for different concentrations of  different concentrations of NO for short‐sensor response 

b li i ti ( h t t d l t ) NO in synthetic air mixture term measurements. baseline variation (short‐term and long‐term)  NO in synthetic air mixture. term measurements.( g )

l il iShort term sensor stability: ConclusionConclusionShort‐term sensor stability: ConclusionConclusionvariations in sensor environment (temperature flow rate humidity)h l

variations in sensor environment (temperature, flow rate, humidity)The wireless FENO measurement prototype.

Th d t i t t th i l i t h i th t ddp yp

The proposed sensory system integrates the signal processing techniques that addressLong‐term sensor stability: p p y y g g p g qimportant challenges in the design of the FENO monitoring device with high sensitivity andI d R d S I l iI d R d S I l i

Long term sensor stability:important challenges in the design of the FENO monitoring device with high sensitivity andIntegrated Readout System ImplementationIntegrated Readout System Implementation drift due to chemical reactionsreliabilityIntegrated Readout System ImplementationIntegrated Readout System Implementationdrift due to chemical reactionsreliability.g y pg y pThe implemented and fabricated readout ASIC that interfaces the sensor/sensor array willR i ti diff t The implemented and fabricated readout ASIC that interfaces the sensor/sensor array willResponse variations across different sensors:be integrated in a nitric oxide breath analyzer for monitoring and managing airway diseases

plarge variability of baseline resistance be integrated in a nitric oxide breath analyzer for monitoring and managing airway diseases, large variability of baseline resistance

such as asthmathe drift in the baseline resistance over time at a different rate such as asthma.VDAC provides adaptable current  the drift in the baseline resistance over time at a different rateThe proposed sensory system for monitoring FENO can be interfaced with different

p pthrough a sensor for different The proposed sensory system for monitoring FENO can be interfaced with different

b dd d lb dd d lthrough a sensor for different 

selective sensors and used as a coarse diagnostic tool to enable an early detection and toEmbedded Signal ProcessingEmbedded Signal Processing base resistance. selective sensors and used as a coarse diagnostic tool to enable an early detection and toEmbedded Signal ProcessingEmbedded Signal Processing base resistance. 

direct more complex diagnostic tools where to focus attentiong gg g

ISEN direct more complex diagnostic tools where to focus attention.Goal: to obtain a calibration curve that would be independent of the baseline variations over

ffGoal: to obtain a calibration curve that would be independent of the baseline variations over 

ReferencesReferencesboth short and long time frame ReferencesReferencesboth short and long time frame.IIN

Y Li P G d M S ć ić "A L P Wid D i R R d IC f B h A l S "IN

Y. Lin, P. Gouma and M. Stanaćević, "A Low‐Power Wide‐Dynamic‐Range Readout IC for Breath Analyzer System",From the controlled measurements with constant gas flow we can extract the response and Proc IEEE Int Symp Circuits and Systems (ISCAS'2013) Beijing China May 2013From the controlled measurements, with constant gas flow, we can extract the response and  Proc. IEEE Int. Symp. Circuits and Systems (ISCAS 2013), Beijing, China, May 2013.P G K K l d X Y M St ć ić d L W “N d B th A l f A irelaxation time of the sensor and obtain the calibration plots P. Gouma, K. Kalyanasundaram, X. Yun, M. Stanaćević and L. Wang, “Nanosensor and Breath Analyzer for AmmoniaIDAC

relaxation time of the sensor and obtain the calibration plots. Detection in Exhaled Human Breath” IEEE Sensor Journal vol 10(1) pp 49 53 2010

IDAC

For the dynamic measurement we are investigating how these parameters characterize the Detection in Exhaled Human Breath , IEEE Sensor Journal, vol. 10(1), pp. 49 53, 2010.L W K K l d M St ć ić d P G “N D i f B th A t D t ti ”

For the dynamic measurement, we are investigating how these parameters characterize the L. Wang, K. Kalyanasundaram, M. Stanaćević and P. Gouma, “Nanosensor Device for Breath Acetone Detection”,IDAC provides path for most of theresponse under different flow rates Sensors Letters, vol. 8, 2010.

IDAC provides path for most of the response under different flow rates.  Sensors Letters, vol. 8, 2010.sensor current. 

AcknowledgmentAcknowledgmentI d d l i (ICA) i d f h i f h b li AcknowledgmentAcknowledgmentIndependent component analysis (ICA) is used for the compensation of the baseline  AcknowledgmentAcknowledgmentp p y ( ) pi ti

l k d f h d dvariations.

Block diagram of the integrated readout circuit This work is supported by the National Science Foundation (NSF) grant IIS‐1231761.g g pp y ( ) g