Wearable Biosensors for Personalized Health …...Wearable Sweat Biosensors: System Level...
Transcript of Wearable Biosensors for Personalized Health …...Wearable Sweat Biosensors: System Level...
Wearable Biosensors for Personalized Health Monitoring
Wei Gao
Assistant Professor of Medical EngineeringDivision of Engineering & Applied Science
California Institute of Technology
Wearable Biosensors for Personalized Medicine
Commercial health monitors can mainly trackphysical activities and vital signs
Challenges and opportunities: physiological monitoring at molecular levels
Human Sweat
Electrolytes• Na+, Cl-, K+, NH4+, Ca2+, H+
Metabolites• Lactate, glucose, urea, uric acid,creatinine
Small Molecules• Amino acids, cortisol, DHEA
Proteins & Peptides• Interleukins, tumor necrosis factor,neuropeptides
Xenobiotics• Heavy metals such as Cu, Hg, Cd,Zn, Pb, As, Ni
• Ethanol• Drugs• Cosmetics
The Current Healthcare Applications of Sweat Test
Current sweat test cannot provide real time information and requires extensive laboratory analysis
Wearable Sweat Analysis
Nature electronics, 2018
Real time, non-invasive, continuous health monitoring
Fully Integrated Wearable Sensors or Perspiration Analysis
Gao et al. Nature, 2016, 529, 509.
• Real time in situ monitoring:• Metabolites (glucose, lactate)• Electrolytes (Na+,K+)• Skin temperature.
• On site signal conditioning,processing, wireless transmission.
• Real time sensor reading calibration.
• Data display on cell phone.
• Data aggregation on cloud server.
‘Smart Wristband’
wireless flexible PCB
flexible sensor array
Wearable Sweat Biosensors: System Level Integration
The platform consists of disposable sensor patch and reusable flexible printed circuit board.
Gao et al. Nature, 2016, 529, 509.
The selection of targeted biomarkers• Metabolites (glucose, lactate)• Electrolytes (Na+,K+)• Skin temperature.
System Level Technology Development - I Flexible Sensors – Enzymatic Sensors (Glucose and Lactate Sensing)
Lactate Sensors
Reference & Counter
CNT+Chitosan
Au
Ag/AgCl
Prussian Blue
CNT+ChitosanLactate Oxidase
e-
Electrode
PB (OX)PB (Red)
GOx (OX)GOx (Red)
Gluconic AcidGlucose
Electrode
PB (OX)PB (Red)
LOx (OX)LOx (Red)
PyruvateLactate
Glucose Sensors
AuPrussian Blue
GOx+CNT+Chitosan
e-
PETPET
Prussian Blue is used as mediator to lower the operation potential(from 0.6~0.7 V to ~0 V) and minimize the interferences.
System Level Technology Development - I Flexible Sensors – Ion Selective Sensors (Na+ and K+ Sensing)
Valinomycin - K ionophore
Metal ion is encapsulated in a molecular cavity whose size is matched to the ion size
According to Nernst Equation:
charge of ion Ion activity/concentration
Na ISE
K ISE
RPVB
Na ionophore X+NaTFPB+DOS+PVCPEDOT:PSS
Au
Valinomycin+NaTPB+DOS+PVCPEDOT:PSS
Au
PVB+NaCl+CNT HydrophobicAg/AgCl
System Level Technology Development - IIFlexible Printed Circuit Board
ADC (built-in)
Phone User Interface
Differential Amplifier (6)
Voltage Buffer (5)
Differential Amplifier (6)
Voltage Buffer (8)
Low-Pass Filter (7)
Low-Pass Filter (9)
Microcontroller (10)
Cloud Servers
Transimpedance Amplifier (1)
Transimpedance Amplifier (3)
Inverter (1) Inverter (3)
Low-Pass Filter (2)
Low-Pass Filter (4)
Voltage Divider
Bluetooth Transceiver (11)
IGlucose ILactate RTemperature VSodium VRef VPotassiumAg/AgCl
Fabrication process flow for the flexible sensors
PETAu
AgAgCl
a Au patterningb
Parylene depositioncO2 plasma etchingd
Ag patterninge Ag chloridationf
gh Flexible arraySensors modification
PET cleaning
Parylene
Nature, 2016, 529, 509.
Characterization of the sensors
All the sensors show linear response vs concentration or logarithm of concentration
0.16
0.18
0.20
0.22
0.24
0.26
0.28
0 15 30 45 60 75 90
Pote
ntia
l (V)
Time (s)
32 mM
16 mM
8 mM
4 mM
2 mM
1 mM
150200250300
0 0.5 1 1.5 2Pote
ntia
l(m
V)
log10([K+], (mM))
[K+]
0.15
0.2
0.25
0.3
0 0.5 1 1.5 2
Pot
entia
l (V
)
log10([K+], (mM))
0.24
0.26
0.28
0.30
0.32
0.34
0 15 30 45 60 75
Pote
ntia
l (V)
Time (s)
160 mM
80 mM
40 mM
20 mM
10 mM[Na+]
210260310360
1 1.5 2 2.5Pot
entia
l(m
V)
log10([Na+], (mM))
0.21
0.26
0.31
0.36
1 1.5 2 2.5
Pot
entia
l (V
)
log10([Na+], (mM))
-6.0
-5.0
-4.0
-3.0
-2.0
-1.0
0 30 60 90 120 150
Cur
rent
(μA)
Time (s)
-6-4-2
0 10 20 30Cur
rent
(µA
)
[lactate] (mM)
[lactate]
5 mM
10 mM
15 mM
20 mM
25 mM
-600
-500
-400
-300
-200
-100
0
100
0 60 120 180 240 300
Cur
rent
(nA)
Time (s)
-600-400-200
0
0 100 200 300Cur
rent
(nA
)
[glucose] (µM)
[glucose]
100 μM
150 μM
200 μM
50 μM
0 μM
System Level Interference & Temperature Compensation
Gao et al. Nature, 2016, 529, 509.
Temperature dependence
a c d b
e g h f
-600
-500
-400
-300
-200
-100
0
100
0 60 120 180 240 300
Cur
rent
(nA
)
Time (s)
-600 -400 -200
0
0 100 200 300 Cur
rent
(nA
)
[glucose] (µM)
[glucose]
100 µM
150 µM
200 µM
50 µM
0 µM
-6.0
-5.0
-4.0
-3.0
-2.0
-1.0
0 30 60 90 120 150
Cur
rent
(µA
)
Time (s)
-6
-4
-2
0 10 20 30 Cur
rent
(µA
)
[lactate] (mM)
[lactate]
5 mM
10 mM
15 mM
20 mM
25 mM
1.90
1.92
1.94
1.96
1.98
2.00
2.02
0 60 120 180 240 300
Res
ista
nce
(kΩ
)
Time (s)
1.90
1.95
2.00
2.05
20 30 40 50 Res
ista
nce
(kΩ
)
Temperature (°C )
temperature 20 °C
25 °C
30 °C
35 °C
40 °C
50
150
250
350
450
550
650
0 240 480 720 960 1,200
Out
put o
f sen
sors
(mV
)
Time (s)
temperature 22 °C
lactate
5 mM
7.5 mM
10 mM
Na+ 20 mM 40 mM 80 mM
K+ 4 mM 8 mM 16 mM
glucose
40 µM 60 µM
80 µM
1.5
1.9
2.3
2.7
20 24 28 32 36 40
Cur
rent
(µA
)
Temperature (°C)
200
300
400
500
Cur
rent
(nA
) 100 µM glucose
5 mM lactate
50
100
150
200
[glu
cose
] (µM
)
0
2.5
5
7.5
10
0 300 600 900 1,200
[lact
ate]
(mM
)
Time (s)
50
75
100
125
150
175
200
0 200 400 600 800 1000 1200 1400
Glucose(Co
ncen
tra.
on(((
no compensation with T compensation
0"
1"
2"
3"
4"
5"
6"
7"
0" 0.5" 1" 1.5" 2" 2.5" 3" 3.5"
Series1"
22 °C 27 °C 32 °C 37 °C
22 °C 27 °C 32 °C 37 °C
50
75
100
125
150
175
200
0 200 400 600 800 1000 1200 1400
Glucose(Co
ncen
tra.
on(((
no compensation with T compensation
0"
1"
2"
3"
4"
5"
6"
7"
0" 0.5" 1" 1.5" 2" 2.5" 3" 3.5"
Series1"
0.24
0.26
0.28
0.30
0.32
0.34
0 15 30 45 60 75
Pot
entia
l (V
)
Time (s)
160 mM
80 mM
40 mM
20 mM
10 mM
[Na+]
210
260
310
360
1 1.5 2 2.5 Pot
entia
l (m
V)
log10([Na+], (mM))
0.21
0.26
0.31
0.36
1 1.5 2 2.5
Pot
entia
l (V
)
log10([Na+], (mM))
0.16
0.18
0.20
0.22
0.24
0.26
0.28
0 15 30 45 60 75 90
Pot
entia
l (V
)
Time (s)
32 mM
16 mM
8 mM
4 mM
2 mM
1 mM
150
200
250
300
0 0.5 1 1.5 2
Pot
entia
l (m
V)
log10([K+], (mM))
[K+]
0.15
0.2
0.25
0.3
0 0.5 1 1.5 2
Pot
entia
l (V
)
log10([K+], (mM))
Interference study
a c d b
e g h f
-600
-500
-400
-300
-200
-100
0
100
0 60 120 180 240 300
Cur
rent
(nA
)
Time (s)
-600 -400 -200
0
0 100 200 300 Cur
rent
(nA
)
[glucose] (µM)
[glucose]
100 µM
150 µM
200 µM
50 µM
0 µM
-6.0
-5.0
-4.0
-3.0
-2.0
-1.0
0 30 60 90 120 150
Cur
rent
(µA
)
Time (s)
-6
-4
-2
0 10 20 30 Cur
rent
(µA
)
[lactate] (mM)
[lactate]
5 mM
10 mM
15 mM
20 mM
25 mM
1.90
1.92
1.94
1.96
1.98
2.00
2.02
0 60 120 180 240 300
Res
ista
nce
(kΩ
)
Time (s)
1.90
1.95
2.00
2.05
20 30 40 50 Res
ista
nce
(kΩ
)
Temperature (°C )
temperature 20 °C
25 °C
30 °C
35 °C
40 °C
50
150
250
350
450
550
650
0 240 480 720 960 1,200
Out
put o
f sen
sors
(mV
)
Time (s)
temperature 22 °C
lactate
5 mM
7.5 mM
10 mM
Na+ 20 mM 40 mM 80 mM
K+ 4 mM 8 mM 16 mM
glucose
40 µM 60 µM
80 µM
1.5
1.9
2.3
2.7
20 24 28 32 36 40 C
urre
nt (µ
A)
Temperature (°C)
200
300
400
500
Cur
rent
(nA
) 100 µM glucose
5 mM lactate
50
100
150
200
[glu
cose
] (µM
)
0
2.5
5
7.5
10
0 300 600 900 1,200
[lact
ate]
(mM
)
Time (s)
50
75
100
125
150
175
200
0 200 400 600 800 1000 1200 1400
Glucose(Co
ncen
tra.
on(((
no compensation with T compensation
0"
1"
2"
3"
4"
5"
6"
7"
0" 0.5" 1" 1.5" 2" 2.5" 3" 3.5"
Series1"
22 °C 27 °C 32 °C 37 °C
22 °C 27 °C 32 °C 37 °C
50
75
100
125
150
175
200
0 200 400 600 800 1000 1200 1400
Glucose(Co
ncen
tra.
on(((
no compensation with T compensation
0"
1"
2"
3"
4"
5"
6"
7"
0" 0.5" 1" 1.5" 2" 2.5" 3" 3.5"
Series1"
0.24
0.26
0.28
0.30
0.32
0.34
0 15 30 45 60 75
Pot
entia
l (V
)
Time (s)
160 mM
80 mM
40 mM
20 mM
10 mM
[Na+]
210
260
310
360
1 1.5 2 2.5 Pot
entia
l (m
V)
log10([Na+], (mM))
0.21
0.26
0.31
0.36
1 1.5 2 2.5
Pot
entia
l (V
)
log10([Na+], (mM))
0.16
0.18
0.20
0.22
0.24
0.26
0.28
0 15 30 45 60 75 90
Pot
entia
l (V
)
Time (s)
32 mM
16 mM
8 mM
4 mM
2 mM
1 mM
150
200
250
300
0 0.5 1 1.5 2
Pot
entia
l (m
V)
log10([K+], (mM))
[K+]
0.15
0.2
0.25
0.3
0 0.5 1 1.5 2
Pot
entia
l (V
)
log10([K+], (mM))
Interference Study
The chemical sensors have good selectivity. Real time temperature compensation is necessary.
Repeatability, Stability and Calibration of the Sensors
For Na and K sensors, one-point calibration is needed.
Na+ and K+ sensors: 1% relative standard deviation in sensitivity Glucose and lactate sensors: 5% relative standard deviation in sensitivity
Na+ Sensors K+ Sensors20
30
40
50
60
70
80
-1 0 1 2 3 4 5
Sen
sitiv
ity (m
v/de
c)
Weeks
20
30
40
50
60
70
80
-1 0 1 2 3 4 5
Sen
sitiv
ity (m
V/d
ec)
Weeks
-400
-350
-300
-250
-200
-150
-100
-1 0 1 2 3 4 5C
urre
nt (n
A)
Weeks
-3
-2.6
-2.2
-1.8
-1.4
-1
-1 0 1 2 3 4 5
Cur
rent
(μA
)
Weeks
Na+ Sensors K+ Sensors
Glucose Sensors Lactate Sensors Glucose Sensors Lactate Sensors-4
-3.5
-3
-2.5
-2
-1.5
-1
0 5 10 15 20
Cur
rent
(μA
)
[Lactate] (mM)
-400
-300
-200
-100
0
0 50 100 150 200
Cur
rent
(nA
)
[Glucose] (μM)
150
200
250
300
350
400
0 1 2 3 4
Pot
entia
l (m
V)
Log10([Na+], (mM))
100
150
200
250
300
350
-1 0 1 2 3
Pot
entia
l (m
V)
Log10([K+], (mM))
Sensor-sensor variations Long term stability
Real time on body sweat analysis
Gao et al. Nature, 2016, 529, 509.
a b c d
‘Smart Wristband’ and ‘Smart Headband’ On Body Validation using Collected Sweat
The Custom-Developed Mobile App
0306090
120
900 1,200 1,500 1,800 2,100 2,400
[Glu
cose
] (μ
M)
Time (s)
20
50
80
110
900 1,200 1,500 1,800 2,100 2,400
[Na+
] (m
M)
Time (s)
ex-situon-body
ex-situon-body
Real time multiplexed sweat analysis during indoor cycling
Gao et al. Nature, 2016, 529, 509.
33 34 35 36 37
T (°
C)
2
5
8
11
[L a
ctat
e]
(mM
)
0 30 60 90
[Glu
cose
] (µ
M)
20
50
80
110
[Na+
] (m
M)
2 5 8
11 14
0 600 1,200 1,800 2,400 3,000
[K+ ]
(mM
)
Time (s)
0
100
200
300
Power(W)
The device can be used to measure detailed sweat profiles and to collect big data.
Sweat starts
Example Application: Dehydration Monitoring toward Sport Medicine
Gao et al. Nature, 2016, 529, 509.
20
30
40
50
60
70
80
0 1,000 2,000 3,000 4,000 5,000 6,000 7,000
[Na+ ]
(mM
)
Time (s)
20
30
40
50
60
70
80
0 1,000 2,000 3,000 4,000 5,000 6,000 7,000
[Na+ ]
(mM
)
Time (s)
dehydration
dehydration
10
20
30
40
50
60
70
80
0 1000 2000 3000 4000 5000 6000 7000
Na+
(mM
)
Time (s)
Subject 1; Water
Subject 2; Water
subject 1; water intakesubject 2; water intake
10
20
30
40
50
60
70
80
0 1000 2000 3000 4000 5000 6000 7000
Na+
(mM
)
Time (s)
Subject 1; Water
Subject 2; Water
subject 3; no water intakesubject 4; no water intake
Sweat sodium can potentially serve as a biomarker for dehydration monitoring.
Wearable Sensors for Ca2+ and pH Monitoring
Nyein, Gao et al. ACS Nano, 2016, 10, 7216.
Kidney function monitoring
Simultaneous monitoring of Ca and pH is essential for accurate Ca analysis.
30
32
34
T (�
C)
Serie…
5.5
6.5
7.5
pH
On-BodypHsensorCommercialpHmeter
00.751.5
2.25
0 5 10 15 20 25 30
[Ca2
+ ] (m
M)
Time (min)
On-bodyCa2+sensorICP-MS
050
100150200
Pow
er O
utpu
t (W
)
pH sensorCommercial pH meter
Ca2+ sensorICP-MS
Temperature
pH selectiveelectrode
Ca2+ selectiveelectrode Temperature
sensor
To FPCB
PVB Reference
Flexible PET
PVB+NaClAg/AgCl Polyaniline
Au
Ca2+H+
Ca2+ sensor pH sensorReference
AuPEDOT:PSS
AuPET PET PET
ETH1001
Ca2+ level in body fluids is dependent on pH
Wearable Sensors for Heavy Metal Monitoring
Gao et al. ACS Sensors, 2016, 1, 866.
Anodic stripping voltammetry for trace level heavy metal analysis
Heavy metal levels in body fluids are extremely low.
Characterization of the Microsensor ArraysAu for Pb, Cu, Hg detection Bi for Zn, Cd, Pb detection
Gao et al. ACS Sensors, 2016, 1, 866.
The microsensor array (Au and Bi) can selectively detect 5 heavy metals.
Heavy Metal Monitoring of Body Fluids
SampleMeasurements from the microsensors (μg/L)
Measurements from ICP-MS (μg/L)
Sweat Cu 249 267Sweat Zn 274 290Urine Cu 571 601Urine Zn 624 634
Zinc detection in human sweat Zinc detection in human urine
Gao et al. ACS Sensors, 2016, 1, 866.
The wearable sensors can accurately measure heavy metals in body fluids.
How to Access Sweat Sample Without Exercise?
Beyond physical exercise: iontophoresis based sweat extraction
Sweat can be induced on demand through iontophoresis.
A Wearable Platform for Sweat Extraction & Sensing
a b
Current
Mode 1: Iontophoresis Mode 2: Sensing
Agonist agent sweat
Sensors CathodeAnodeHydrogelHydrogel
Sensor 1 Sensor 2I1 or V1 I2 or V2
Bluetooth transceiver
Phone user interface
Microcontroller
DAC ADC
Low-pass filter
Analog frontend
Sensor A
Low-pass filter
Analog frontend
Sensor B
Current source
Protection circuit
Electrodes/Hydrogel
Mode 1: Iontophoresis Mode 2: Sensing
c
d
Skin Skin
Na ISE
Iontophoresis electrodes
Cl ISE
PVB Reference1 cm
Beyond physical exercise: accessing sweat samples on demand using iontophoresis
Current
Mode 1: Iontophoresis
Mode 2: Sensing
Agonist agent
sweat
Sensors CathodeAnodeHydrogelHydrogel
Sensor 1 Sensor 2I1 or V1 I2 or V2
PNAS, 2017, 114, 4625
Iontophoresis based Sweat Extraction
97
35
284
7.25
1
52 45
354
5.25
9.5
389
60
230.
8
24 35
162
60
252.
4
20
35
111
60
198
23 20
240
60
210
39 30
97
35
284
7.25
1
52 45
354
5.25 9.5
389
60
230.
8
24
35
162
60
252.
4
20
35
111
60
198
23 20
240
60
210
39
30
R ESPON SE LATEN C Y
( SEC ON D S)
R ESPON SE D U R AT ION ( M IN U TES)
PEAK SW EAT R ATE
( N L/M IN /C M 2)
T IM E TO PEAK ( M IN U TES)
T IM E AT PEAK R ATE ( M IN U TES)
Acetylcholine 1% Acetylcholine 10% Methacholine 1%Methacholine 10% Pilocarpine 0.5% Pilocarpine 5%
97
35
284
7.25
1
52 45
354
5.25 9.5
389
60
230.
8
24
35
162
60
252.
4
20
35
111
60
198
23 20
240
60
210
39
30
R ESPON SE LATEN C Y
( SEC ON D S)
R ESPON SE D U R AT ION ( M IN U TES)
PEAK SW EAT R ATE
( N L/M IN /C M 2)
T IM E TO PEAK ( M IN U TES)
T IM E AT PEAK R ATE ( M IN U TES)
Acetylcholine 1% Acetylcholine 10% Methacholine 1%Methacholine 10% Pilocarpine 0.5% Pilocarpine 5%
97
35
284
7.25
1
52 45
354
5.25 9.5
389
60
230.
8
24
35
162
60
252.
4
20
35
111
60
198
23 20
240
60
210
39
30
R ESPON SE LATEN C Y
( SEC ON D S)
R ESPON SE D U R AT ION ( M IN U TES)
PEAK SW EAT R ATE
( N L/M IN /C M 2)
T IM E TO PEAK ( M IN U TES)
T IM E AT PEAK R ATE ( M IN U TES)
Acetylcholine 1% Acetylcholine 10% Methacholine 1%Methacholine 10% Pilocarpine 0.5% Pilocarpine 5%
97
35
284
7.25
1
52 45
354
5.25 9.5
389
60
230.
8
24
35
162
60
252.
4
20
35
111
60
198
23 20
240
60
210
39
30
R ESPON SE LATEN C Y
( SEC ON D S)
R ESPON SE D U R AT ION ( M IN U TES)
PEAK SW EAT R AT E
( N L/M IN /C M 2)
T IM E TO PEAK ( M IN U T ES)
T IM E AT PEAK R ATE ( M IN U TES)
Acetylcholine 1% Acetylcholine 10% Methacholine 1%Methacholine 10% Pilocarpine 0.5% Pilocarpine 5%
97
35
284
7.25
1
52 45
354
5.25 9.5
389
60
230.
8
24
35
162
60
252.
4
20
35
111
60
198
23 20
240
60
210
39
30
R ESPON SE LATEN C Y
( SEC ON D S)
R ESPON SE D U R AT ION ( M IN U TES)
PEAK SW EAT R ATE
( N L/M IN /C M 2)
T IM E TO PEAK ( M IN U TES)
T IM E AT PEAK R ATE ( M IN U TES)
Acetylcholine 1% Acetylcholine 10% Methacholine 1%Methacholine 10% Pilocarpine 0.5% Pilocarpine 5%
97
35
284
7.25
1
52 45
354
5.25 9.5
389
60
230.
8
24
35
162
60
252.
4
20
35
111
60
198
23 20
240
60
210
39
30
R ESPON SE LATEN C Y
( SEC ON D S)
R ESPON SE D U R AT ION ( M IN U TES)
PEAK SW EAT R ATE
( N L/M IN /C M 2)
T IM E TO PEAK ( M IN U TES)
T IM E AT PEAK R ATE ( M IN U TES)
Acetylcholine 1% Acetylcholine 10% Methacholine 1%Methacholine 10% Pilocarpine 0.5% Pilocarpine 5%
Pilocarpine 0.5%Pilocarpine 5%
Methacholine 1%Methacholine 10%Acetylcholine 10%
Acetylcholine 1%
ResponseLatency (s)
ResponseDuration (min)
Time toPeak (min)
Peak SweatRate
(nL/min/cm2)
Time atPeak rate
(min)
Sweat extraction can be controlled by type of drugs and the drug dosage. PNAS, 2017, 114, 4625
Periodical Sweat Extraction using the Wearable Platform
Sweat can be induced periodically.
10 s
300 s
PNAS, 2017, 114, 4625
Example Applications of Wearable Sweat Biosensors
Non-Invasive Glucose MonitoringMedical monitoring and diagnosis without accessing blood
0
30
60
90
120
150
Con
cent
ratio
n(m
M)
Cl- Na+ Na+Cl-
Healthy subjects
Cystic FibrosisPatients
Swea
tglu
cose
(nA)
0
40
80
120
160
Bloo
dgl
ucos
e(m
g/dL
) Fasting After glucose intake
Blood glucose Sweat glucose20
30
40
50
60
70
80
0 1500 3000 4500 6000 7500
[Na+
] (m
M)
Time (s)
dehydration
dehydration10
20
30
40
50
60
70
80
0 1000 2000 3000 4000 5000 6000 7000
Na+ (
mM
)
Time (s)
Subject 1; Water
Subject 2; Water
subject 1subject 2
Wearable sweat sensors enable the correlation studies between sweat biomarkers and blood biomarkers
PNAS, 2017, 114, 4625
Example applications of wearable sweat biosensors
Cystic Fibrosis Diagnosis
Medical monitoring and diagnosis without accessing blood
PNAS, 2017, 114, 4625
Example applications of wearable sweat biosensors
Cystic Fibrosis Diagnosis
Medical monitoring and diagnosis without accessing blood
Wearable sweat sensors can be used for Cystic Fibrosis screening and diagnosis.
0
30
60
90
120
150
Con
cent
ratio
n(m
M)
Cl- Na+ Na+Cl-
Healthy subjects
Cystic FibrosisPatients
Swea
tglu
cose
(nA)
0
40
80
120
160
Bloo
dgl
ucos
e(m
g/dL
) Fasting After glucose intake
Blood glucose Sweat glucose20
30
40
50
60
70
80
0 1500 3000 4500 6000 7500
[Na+
] (m
M)
Time (s)
dehydration
dehydration10
20
30
40
50
60
70
80
0 1000 2000 3000 4000 5000 6000 7000
Na+ (
mM
)
Time (s)
Subject 1; Water
Subject 2; Water
subject 1subject 2
PNAS, 2017, 114, 4625
Drug Monitoring with Wearable Sweat Sensors
Adv. Mater. 2018, 1707442
Methylxanthine Drug Monitoring with Wearable Sweat Sensors
Adv. Mater. 2018, 1707442
Characterization of caffeine sensor
Sweat caffeine vs caffeine dosage (through iontophoresis sweat)
Methylxanthine Drug Monitoring with Wearable Sweat Sensors
Adv. Mater. 2018, 1707442
Dynamic drug metabolism monitoring
Promising use in clinical pharmacology and precision medicine, such as therapeutic drug monitoring, drug abuse intervention, and other aspects of the drug-related healthcare system.
Microfluidic based Sweat Sampling and Sweat Rate Monitoring
ACS Sensors 2018
Sweat composition concentrations and sweat rate are inextricably linked.
Roll-to-Roll Gravure Printing – Mass Production of Sensors at Ultralow Cost
ACS Nano 2018, 10.1021/acsnano.8b02505
This development of robust and versatile R2R gravure printed electrodes represents a key translational step in enabling large-scale, low-cost fabrication of disposable wearable sensors for personalized medicine.
Other promising applications
Exp. Physiol. 84, 401–404 (1999).
Biol Psychiatry. 2008, 64, 907-911
Metabolomics (2016) 12:166
Lung cancer screening
Major depressive disordersSubstance abuse
Nature Electronics, 2018
Bio-Integrated Lab on the Skin
Wearable/point-of-care devices
• Body fluid analysis – sweat, ISF, plasma, saliva
• Label free protein/hormone analysis
• EEG• ECG• Blood pulse pressure (HR, HRV)• Skin temperature
Machine Learning
Vital sign sensors
Molecular sensors
Thank you for your attention!
Questions?