• No results found

Communicating sensors for better diagnosis

N/A
N/A
Protected

Academic year: 2022

Share "Communicating sensors for better diagnosis"

Copied!
19
0
0

Loading.... (view fulltext now)

Full text

(1)

Communicating sensors for better diagnosis

Prof. Jerker Delsing

Luleå University of Technology, EISLAB

(2)

Medical measurements

 Pulse

 Blood pressure

 ECG

 EMG

 Lung sound

 Breathing rate

 ...

 ..

 .

 .

 .

(3)

Measurements conditions

 Static

Patient at hospital

Present state of the art

 Dynamic

Patient in real living situation

Possible under some circumstances

(4)

Barriers to dynamic measurements

 Limited understanding of dynamic data

 How to analyze data automatically

 Non feasible technology

 Sensors are

 to large - to heavy - to expensive!

 to difficult to apply properly!

 Having to short energy life time!

 How to communicated data in a secure manner!

 How to integrate sensor into health care application

 How to make field calibration?

(5)

Technology development

 Small & cheap electronics

Requires large volume markets to justify development cost

 Energy consumption Rapidly reducing

 Security

What is adequate security?

 System integration

Still along way to go

(6)

Technology development

 Sensor application in real life situations A very long way to go!

 Field calibration

Yet not heard about

(7)

The Embedded Internet System Vision

 Sensors on the Internet

 Minimal size < 1 cm 3

 Power life time > 2 year

 Wireless connection

 TCP/IP and web-services

 Ad-hoc communication

 Ad-hoc system integration

 Secure

 Simple application

 Roughed packaging 1 cm 3

(8)

MULLE – EIS platform - today

 Wireless Internet capable sensor platform

 4 cm

3

including battery

 Life time: Month to years

 5uW reactive awake power consumption

 10 MHz M16, 20kRam 256kFlash

 Bluetooth, ZigBee

 Power harvesting possible

 EIS sensor network functionality – TCP/IP, Web-server

– Ad-hoc networking – Service discovery

– Real time: Timber or RTOS

– Data encryption

(9)

The thumb size ultrasonic measurement system

 Electronics integrated on piezo-ceramics

 Combined with MULLE

(10)

EIS ad-hoc network example

(11)

Examples of trails

(12)

Professor Pulse – Vasaloppet

(13)

Whats next

(14)

ECG

 < 50g

 > 1 week of battery life time

(15)

Fall sensor

 On-line fall sensor with very long battery life time

 On-board data analysis

(16)

Data handling

 Real time communication using GPRS over patient cell phone Cell phone limiting device

 Local storage of data possible G-byte of data possible

 Real time data analysis possible

Requires more but still limited power on-board

(17)

Security

 What is adequate security for dynamic medical data??

Bank level security?

..

..

No security

 Authentication

 Data encryption

 Device tampering

 Data tampering

 Device flooding

 Availability in 112 situations

(18)

What is possible in 4 cm 3

 Authentication

 Data encryption

 Secure link

 ..

 ..

 Game of resources Power

Communication bandwidth

(19)

Conclusions

 Mobile sensors for dynamic health monitoring is emerging

 Dynamic data analysis has to improve

 Possibilities for on-line data analysis exist but has to improve

 On-line data analysis need improvement and acceptance

 Feasible security is possible

 Easy application is still a major hurdle

References

Related documents

Figure 4: Schematics of sample dilution. The red lines represent the position of the sample at certain times. The sampling pump starts a set amount of time before the valve

Within this category we tried to identify behaviors that can be seen as stereotypical, such as expressions of hesitation, fear or shyness for female characters versus leader- ship

Thus issues like sensor size, weight, means of communication, field calibration, power lifetime etc.. becomes

After a discussion of the main challenges in microsystems con- troller design, it has been shown that a digit-serial processing scheme based on on-line arithmetic is particularly

If measurements of leading vehicles also are available from the radar tracker (tracks) the quality of the estimate can be improved a lot and can handle sig- nificantly longer

The Java Native Interface (JNI) is a programming framework that allows Java code running in the Java virtual machine (JVM) to call and be called by native applications

(b) Image of the resulting point cloud from registering sequence two using the extension of removing low keypoint match scoring point clouds.. Figure 5.4: Sequence 2: Figure 5.4a

• When rolling n-day averages of rates of change of time series data are synthesised, how do the methods of virtual, dynamic and materialized views compare in terms of total