Upload
others
View
4
Download
0
Embed Size (px)
Citation preview
S h e n z h e n , C h i n a , 2 2 n d D e c . 2 0 1 7
Wearable Body Sensor Networks
L e i W a n g , r e s e a r c h p r o f e s s o r , d e p u t y d i r e c t o r
I n s t i t u t e o f B i o m e d i c a l a n d H e a l t h E n g i n e e r i n g
S h e n z h e n i n s t i t u t e s o f A d v a n c e d T e c h n o l o g y
C h i n e s e A c a d e m y o f S c i e n c e s
OUTLOOK
Proprietary proof-of-concept systems
Technology evolution
Part 2
Part 1
Academic contributionPart 3
Translational effortPart 4
Ongoing research activitiesPart 5
PART 1 Technology evolution
OUTLOOK
1.1 Technical history – The beginning
The Wearable concept started on 1980s.
Wearable computing is the study or practice of inventing,
designing, building, or using miniature body-borne (i.e.
“Smart Clothing” (Mann, 1996a)).
Bearable Computing.
Constancy of interaction, that the human and computer
are inextricably intertwined.
Muti-task 1991: Started Wearable Computing project
at MIT.
1995 : World’s first convert wearable
computer – camera and display concealed
and in ordinary eyeglasses .
The father of Wearable Computing
Representative work
ICON: prof. Steven Mann
Thad Starner’s system Wearable Wireless webcam
Prototyping setup
Bluetooth transceiver kit
1.2 Technical history – The late 1990s and earlier 21st century
Millennium: wearable monitoring devices with focuses on
ASIC design
Wearable electronics
Sensium platform, and others around world
Vital Jacket (Biodevices).
Disposable electrodes embedded in T-shirt.
Bluetooth transmission, software.
Low-power wireless sensor interface platform.
Single-chip wireless
ICON: prof. Yong Lian
Representative work
Flexible wireless ECG sensor
Definition of ECG
ECG refers to the heart in each cardiac cycle, by the pacemaker, atrium, ventricle have been excited,
accompanied by changes in bioelectricity, through the skin of the electrode from the body surface leads to a
variety of forms of potential changes in the graphics.
ECG is the heart of the excitement of the occurrence, dissemination and recovery process of objective
indicators, where feel chest tightness, palpitations, palpitation, dizziness, vertigo, precordial discomfort or pain
and other symptoms are required ECG examination.
Value of ECG
Arrhythmia
Heart rate variation
Myocardial ischemia
Myocardial infarction
Conduction time
Cardiac output
1.2 many ECG monitoring prototypes
The Importance of Dynamic Continuous ECG Recording on Early Warning of Cardiovascular and
Cerebrovascular Events as an Example
Dynamic ECG monitoring is very important for the early detection of heart
disease, early treatment and early warning prevention;
Dynamic ECG monitoring can greatly improve the positive rate of heart
disease, For the heart physician to develop or adjust the treatment
strategy to help, Emergency alerts instrument can reduce the chance of
sudden cardiac death;
Dynamic ECG monitoring can provide health care data for medical
services, insurance companies and pharmaceutical companies to
improve the accuracy of health services.
ST-T severe ECG is an important indicator of
early coronary heart disease diagnosis;
Heart rate variability is an important factor in
predicting the occurrence of sudden death;
Reduce the risk of arrhythmia;
Evaluation of drug treatment tracking.
1.2 many ECG monitoring prototypes
1.3 Technical history – BSN earlier 21st century
Advances in wireless communication, sensor design, and
energy storage technologies have meant that the concept
of a truly pervasive (WSN) is rapidly becoming a reality.
Integrated micro sensors no more than a few millimeters in
size, with onboard processing and wireless data transfer
capability are the basic components.
Director of the Hamlyn Centre,
Imperial College London, UK
FREng IEEE Fellow
Representative work
ICON: prof. Guang Zhong
Yang
skeletal muscles
central nervous
peripheral nervous
rigid link segment
movement
external forces
1.3 Motion motion motion
Body Sensor Networks Symposium 2004
14th 13th
1.3 Body sensor network (BSN) conferences 2004-2018
1.4 Technical history – Wearable medical devices 2010s
Personal wearable medical devices and sensors have
become more and more prevalent and popular.
The development of BSN for medical applications
Standardized low-power wireless network with
communication protocols for BSN and clinically approved
wearable devices to be used as nodes for BSN.
Using wearable medical devices and sensors as the nods
of body sensor networks (BSN) could allow better long-
term monitoring of health condition.
Representative work
ICON: prof. Yuan
Ting Zhang
Research Interests:
Wearable medical devices
Unobtrusive physiological measurments
Body sensor networks
Physiological modeling
Cardiovascular health informatics
1.4 Current MEDICAL wearable devices
Wearable“TENS Transcutaneous electrical
nerve stimulation” Anti-pain device
Wearable insulin pump
– artificial pancreas
Wearable exoskeleton
- artificial muscle
Wearable hemodialysis
- artificial kidney
Wearable rehabilitation
functional stimulation shirtWearable shirt for atrial defibrillation
Wearable miniature
electrocardiogram
In recent years , Food and Drug Administration (FDA) approved a variety of wearable
medical equipment, were approved for medical use.
Artificial auxiliary organs Electrophysiological diagnosis
Functional electric stimulation
Go flexible
Chemical sensing devices made from flexible electrode
materials could be incorporated into clothing or attached
directly to the body .
Continuous physical – chemical - biological monitoring.
E-textile
Wearable robot (not only sensor, but also actuation)
Stretchable silicon
integrated circuit Skin-like devices for
blood oximetry
Energy harvester on the surface of heartE-Skin Tattoo BiosensorsFlexible Sensors for On-body Sensing
1.5 Wearable Devices – Present time
Representative work
ICON: prof. Joseph
Wang
1.5 Medical health is a key of area wearable devices- market expectations
The rapid warming of the wearable device market
has attracted many enterprises, manufacturers and
consumers. For now, the market is in the initial stage
and is waiting for the leader to appear.
China's wearable device sales have more than 10 billion ¥, reaching 11.49 billion, of which
70% involved in the field of medical health in 2016.
25 companies investing in the global wearable device market
Wearable technology is One of the ten cutting-edge technologies that affect the future of
mankind(World Economic Forum,2015)
glucose testing
Nature 531(7596), 2016
Electromagnetic
regulation
Science 345(6200),2014
Flexible substrate
Science 355(59), 2017
diagnostic imaging
study
Nature 542(542),2017
nanoporous long
nucleic acids
Nature 530(7589),2016
Bionic photomicrographs
Nature ,531(7594),2016
Bionic system
Science 356(1280), 2017
pressure monitoring
Nature, 527(456),2015
Photosensitive material Multi-parameter sensor
Nature, 529(7587) ,2016
1.6 Wearable Body sensor network – near future new surge
Nature, 546(632),2017
smart fabric artificial muscle
Science 343(6173),2014
Flexible skin
Science 357(6353) ,2017
1.7 short summary
Physiological parameter detection: Develop new materials such as electrodes,
New biosensors and electronic fabrics, Core components of wearable
biosensor;
Wearable device integration: Bionic micro-nano fabircatiion, Low power IC
design and High density packaging, Flexible microsystems;
In-body data communication: Human body near-filed efficient communication
principle and safety theory;
Health status classification and prediction: Statistical learning method for
wearable data sets and Chronic disease risk prediction;
Wearable system energy supply: Nano-generator, Flexible light energy material,
Energy harvesting method using temperature gradient and Develop principle
prototype is developed;
Medical and health applications: Explore the integration of wearable medical
equipment and fashion, innovative design, Explore industry, group standards
and business models, Expand wearable robots and other new applications.
OUTLOOK
PART 2 Proprietary proof-of-concept systems
the first developed single-point flexible ECG acquisition system
Design patterned gold electrodes
Design of micro-pillared structures attachment layer
Configurable IC
Miniaturized hardware acquisition system
Design of a single point composite structure
2.1 ECG Patch system
the first developed single-point flexible ECG acquisition system
Design patterned gold electrodes
Design of micro-pillared structures attachment layer
Configurable IC
Miniaturized hardware acquisition system
Design of a single point composite structure
2.2 EEG glasses frame
Die:1.456mm*0.83mm
2.3 Analogue Front-end IC
YN
Low Noise, Low Frequency, Low Power ASIC
Parameter Value
Acquisition channels 2/4/8
Power consumption 80 uW/Chl
Input referred noise 1.5 uVrms
Programmable gain 40-8000
Bandwidth range 0.05~200 Hz
CMRR ~100 dB
BlocksPreamplifiers, filters, gain
amplifiers, clocks, reference, LDOs, SPI…
Package QFN32, QFN24,LQFP64
♣ Noise modulation Low noise
♣ Current-steering filter Low frequency
♣ Log domain circuit Low power
♣ Global CM feedback High CMRR
2.4 Fall detection
Predicted the fall time, the critical angle and developed one new method of fall warning
based on the BSN location optimization and parameter optimization
2.5 Human body communication
To achieve the transmission rate of 2Mbps human communication in vivo
experiments, bit error rate 1e-6
1
2
3
4
5
3. Patch heart rate meter
4. EEG monitoring glasses
1. Human body communication demonstration system
2. Motion sensing belt5. Temperature watch
2.6 other prototypes10-s video
2.7 Technology “Know how”
Flexible wearable core IC design, MEMS integration and e-Textile
technology development
Wearable human body communication technology research and
development and application of high-tech strategy
Wearable movement perception and its application in screening for
vulnerable population and sports rehabilitation evaluation
Mini ECG, brain electric eye frame, fall prevention belts, watches and other
wearable health new equipment development
PART 3 Academic contribution
- Selected SCI papers in 2015-2017
OUTLOOK
TITLE : Technique for Fetal ECG Extraction Using Single Abdominal Recording
Method: Singular value decomposition (SVD) and smooth window (SW)
techniques are combined to build reference signal in an adaptive noise
canceller (ANC) for fetal ECG extraction.
Conclusion: Validation of the proposed method with signals from a public dataset,
and a self-recorded private dataset showed that the proposed method achieved F1
scores of 99.61%, 98.58% respectively for the detection of fetal QRS
abdominal recording
y(n) = f(n) + x(n)
+
-reference
MECGx’(n)x̂(n)
f̂(n)
FECG
reference-formation
scheme
(SWSVD)
adaptive
filter
0 500 1000 1500 2000 2500-60
-40
-20
0
20
40第1路母体腹壁信号
200 400 600 800 1000 1200 1400-50
0
50
100
150第2路母体腹壁信号
0 500 1000 1500 2000 2500-80
-60
-40
-20
0
20
40第3路母体腹壁信号
1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300-50
0
50
100abdominal recording 2
1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300-500
0
500
1000thoracic recording 2
1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300-20
0
20extracted fecg
1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300-50
0
50
100abdominal recording 2
1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300-500
0
500
1000thoracic recording 2
1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300-20
0
20extracted fecg
MECGQRS-detection
smooth
SVD
concatenate
QRS-length (N segments)
No fetal QRS
QRS-extraction
Inter-QRSlength
abdominalrecording
estimated MECG
estimated QRS
estimated Inter-QRS
preprocess
High-Freq cutoff fh Low-Freq cutoff fb
Year:2017, Volume:17
3.1 – Wearable Computing (1/5)
3.1 – Wearable Computing (2/5)
TITLE : An explorative investigation of functional differences in plantar center of
pressure of four foot types using sample entropy method.
Year:2017, Volume:55
Method: sample entropy was used to quantify complexity and regularity of
medial-lateral and anterior-posterior displacements, and the vertical ground reaction
force of the center of pressure during the stance phase.
Conclusion: When investigating foot function, it is important to take into account dynamic
characteristics of the progression of the center of pressure that contain the dynamic information
about walking pattern.
TITLE : Automatic Extraction of Central Tendon of Rectus Femoris (CT-RF) in Ultrasound
Images Using a New Intensity-Compensated Free-Form Deformation-Based Tracking Algorithm
With Local Shape Refinement
Method: Proposed a new IC-FFD-based tracking algorithm with LSR for extracting the shape
deformation of CT-RF.
Conclusion: The proposed algorithm offers better tracking performance and serves as a
valuable tool for automatic quantitative analysis of CT-RF in sports science as well as
rehabilitation assessment.
Year:2017, Volume:21
3.1 – Wearable Computing (3/5)
TITLE : An investigation into the bilateral functional differences of the lower limb
muscles in standing and walking
Method: We investigated whether characteristics of left limb and the one of the right
limb have the same statistical characteristics using Wilcoxon rank-sum test, and studied
dynamic signal irregularity degree for sEMG activities via sample entropy.
Conclusion: At different speeds, active degrees of different muscles were significant
different between left limb and right limb.
Year:2016, Volume:
3.1 – Wearable Computing (4/5)
TITLE : Feature Selection and Predictors of Falls with Foot Force Sensors Using KNN-
Based Algorithms
Method: Describes a new algorithm for classifying fallers and non-fallers using K-
nearest neighbor (KNN)-based classifiers on force platforms.
Conclusion: LMPNN outperform the other algorithms with 100% of accuracy, 100% of
sensitivity and 100% of specificity, and reached the maximum when k was equal to 3.
0 2 4 6 8 10 12 14
0.7
0.8
0.9
1
k
The classification rate of LMPNN rule
Accuracy rate
Sensitivity rate
Specificity rate
0 2 4 6 8 10 12 140.2
0.4
0.6
0.8
1
k
The classification rate of PNN rule
0 2 4 6 8 10 12 140.2
0.4
0.6
0.8
1
k
The classification rate of LMKNN rule
Year:2015, Volume:15
3.1 – Wearable Computing (5/5)
TITLE : Freestanding electrostatic scratch drive microstructures using lamination of
photosensitive films for microfluidics and microrobotics applications
Method: The microstructures are developed using lamination of SU-8 photosensitive
films and deposition of copper (Cu), which has potential for large scale manufacturing.
Conclusion: The microstructures were actuated moving on an electric substrate by
varying the voltage. It was found that the microstructure step size is proportional to
the square of voltage. When 1000 V, 50 Hz electric field was applied, the step size was
0.1 μm and average speed could achieve 300 μm/min, which shows great potential in
microfluidics and microrobotics applications.
Year:2017, Volume:23
3.1 – Wearable Sensor (1/3)
TITLE: Tunable water sensitive polymeric composites with synergistic
graphene and carbon nanotubes
Method: The synergistic rGO-CNTs were found to magnify the water sensing,
which may attribute to the swelling effect on the CNTs entanglement segregated by
the rGO sheets. The findings may greatly benefit the exploration of nanocarbons in
the field of the flexible sensory materials.
论文里面的图片一张 Year:2017, Volume:199
3.1 – Wearable Sensor (2/3)
TITLE: Toward a Smartphone Application for Estimation of Pulse Transit Time
Method: Presents a PTT estimation method based on photoplethysmographic imaging
(PPGi).
Conclusion: The proposed method is especially suitable for implementation in dual-camera-smartphones, which could facilitate PTT measurement among populations affected by cardiac complications.
Year:2015, Volume:15
3.1 – Wearable Sensor (3/3)
TITLE : Characterization of In-body Radio Channels for Wireless Implants
Method: An inhomogeneous human body model was proposed to study the in-body
radio channels (I-BRCs) for wireless implants. In addition, the model was verified by
experimental measurements with a 45kg pig.
Conclusion: The result showed that the gain of I-BRCs at human body communication
(HBC) frequency band was insensitive with electrode sizes and electrode types.
Year:2017, Volume:17, Issue:5
3.1 – Wearable Communication (1/2)
TITLE : An Approach to Biometric Verification Based on Human Body Communication in Wearable Devices
Method: The gain S21 of volunteer’s forearm is acquired by VNA, and a threshold-
adaptive template matching (TATM) algorithm based on weighted Euclidean distance is
proposed for rapid verification in wearable devices.
Conclusion: The equal error rate (EER) based on TATM algorithm is 7.06% at verification
threshold T2=1.91.
Year:2017, Volume:17, Issue:1
3.1 – Wearable Communication (2/2)
To explore the frontiers of human physiology,Promote the integration of life science and
electronic technology, Solve the biomedical and electronic problems in the field of
population and health, and Raise the level of scientific and technological innovation.
Electronics Biomedical Science
Biomedical electronicsChip design
Circuit system
Signal
processing
… …
Biomedical micro -
nanosensor
Biomedical images
Biomedical Informatics
Biomedical SciencePhysics
The human body is a significant application area of the semi-conductor industry.
Institute of Biomedical Engineering, Imperial College.
PART 4 Translational effort
OUTLOOK
4.1 Low-cost Healthcare
One Target
Disciplinary Position: Multidisciplinary
Cross
Type Position:Industrial Technology
Research Institute
Mission Position : Promote new
industries
Low-cost
Healthcare
High-end medical images
Medical rehabilitation robotTarget Postion : World-class level
Three Significant Breakthroughs
Bring routine ECG examinations for Rural clinic via technology innovations
Proprietary IPs – lower thedevice cost
Improve the performance
Design flexibility, various approaches
Data exchange standardizations
Mobile Internet => mobile healthcare
4.1 Low-cost Healthcare
4.2 Technology translation
Participate in ten registered product standards written,
are approved by
Series of products obtained
medical device registration card
Solutions were injected into four companies with a range of primary healthcare equipments
4.3 CASE study Passing National standard YY1079- pain but worth doing
(up)--EGC-1C and ECG-1CMRR
(down)-iBUSS signal source and
test box
YY1079 Specification√Sensitivity
√ Noise level
√Baseline drift
√ Time constant
√ Input impendence
√ Calibration voltage
√ Input circuit current
√ High polarization voltage
√ 50Hz power line interference
√ Common mode rejection ratio
√Amplitude-frequency characteristic
Require tremendous DOMAIN knowledge on Biomedical Engineering
and Microelectronics
2016 the system in village clinic market share is
the first in China which the total share 30%
Cumulative number of the devices, successful bidder (top ten manufacturers)
Passed technical assessment of
Medical Equipment Association
4.5 Highlight - No. 1 market share
Hat
chin
g e
nte
rprise
s g
ot
multip
le r
ound
s of
finan
cing
① Proof-of-Concept
- Scientific Ideas
② Simulation and
Modelling
③ Prototyping
④ Demonstration of the
application – Business Model
⑤ Product Standards
⑥ Manufacture,
Marketing and Sale
Analog and digital IC design, embedded system development,
mechanical design(3D printing), data analysis
SCIENCE
PATENT
TECHNOLOGY
MARKET
CAPITAL
INDUSTRY
OUTLOOK
PART 5 Ongoing research activities
5.1 Modeling and Simulation for Wearable Communication
Ground
out in
电极耦合电极耦合
隔离器隔离器
Modeling based on field-circuitNumerical simulation with finite element
Fabrication of multilayer tissue mimicking phantoms
Electric effect
Thermal effect
Mechanical effect
Chemical effect
Dielectric properties
adjustable
High melting point
Easy to shape
High stability
Anti-corrosion
Weakening electrode
polarization effectMeasurement with
whole-body phantom
National Natural Science Funding Committee – U1505251
5.2EM-based Blood Glucose Monitoring / Personal Identification
in vivo experiment
Inhomogeneous media model
National Defense Innovation project
5.3 Wearable Robot for percutaneous coronary intervention
To improve Haptic To further reduce radiation To promote positioning accuracy
National Natural Science Funding Committee – U171320090
5.4 - Flexible wearable sensors – beyond ExG
Room-temperature liquid metal
based physical sensors
• FEA analysis of stretchable sensors
• Mechanical characterization of the sensors
• Electrical characterization of the sensors
Wearable graphene enhanced
sweat sensor
• Electrochemical characterization of the sensors
• Mechanical characterization of the sensors
Chinese Academy of Sciences – Key project
5.5 Bionic Self-tuning Soft Robot for Knee
49
Bone Movement Soft tissue EMG
Image Processing
Musculoskeletal Simulation
Identify the key muscles of knee affecting gait stability
Bi-planar X ray
Reducing Energy Cost in Biomechanics
Nanshan Innovation Team