Deteksjon av redusert blodtilførsel til hjertet ved hjelp av akselerometerteknologi

Detection of regional cardiac ischemia by accelerometertechnology

Ole Jakob ElleThe Interventional Centre

Goals– Sensor measuring heart motion– Detection of graft occlusion and ischemia

Collaboration– Medical: The interventional Centre, Rikshospitalet– Research: The interventional Centre, Rikshospitalet– Technology: Vestfold University College– Commercial: BMI, Biomedisinsk Innovasjon AS

Contributors– The Interventional Centre : Ole Jakob Elle, Steinar Halvorsen, Harald

Dugstad, Halfdan Ihlen, Tor Inge Tønnessen, Erik Fosse– Vestfold University College : Lars Hoff, Craig Lowrie, Kristin Imenes,

Morten Grimnes, Hans Jørgen Alker

Background

Impaired coronary circulation causes heart muscle hibernation

Immediate reduction in contractile workOccluded coronary graft cause regional

movement reduction

AccelerometersEarly use

– Seismograph Vibrations, earth quakes– Navigation, military Rockets, submarines

1990s– Microtechnology – Small, inexpensive units available

Today: Used everywhere– Automotive: Airbag crash-sensors– Cars, consumer electronics, medical, …

Easily available– Price ~ USD 3– Volume ~ 100 mill. units per year

Cardiac Ischemia change in myocardial function and motion

The idea and hypotesis: – 3-D accelerometers could detect regional systolic changes due to

ischemia or reduced blood flowPatented by The Interventional Centre : WO 03/061473 A1

First prototype 3-axis accelerometer sensor

Recording Signals on Computer

Calculation of Heart Motion• Position

Motion dominated by respirationPatient movements and gravity: Other sources of acceleration

High pass filtering at 1.0 Hz to isolate heart’s own motion

Position from acceleration: Numerical integration

Calculated Heart MotionAcceleration

Remove movements from respiration, patient movements etc.Integrate twice wrt. time

Heart position• Nice curves. Noise hardly visible• Arrhythmias easily identified

Arrhythmias

Prototype of 3 Axis Sensor• Commercial accelerometers –

(Analogue Devices)– Two at 90º angle – Each sensor: 2 axes – Two sensors: 3 axes

• One axis redundant

• Size: 6x6x12 mm• Encapsulated• Sutured to heart wall

The model: Porcine

6 animalsECGhemodynamicsFlow

– aorta, LADAccelerometers

– LAD and CX regionOccluder on LAD distally

to the first diagonal branch(60 sec. LAD occlusion)

Accelerometer A

Accelerometer B

Flowmeter and occluder,

Sound analogy

• Sound is changes in air density due to wavepropagation in air.

• A different tone is heard if the sound frequency is changed

• Sound comprises a set of frequency components(harmonics).

• The heart surface can be seen as a loudspeakermembrane and the accelerometer as a microphone

• Changes in motion pattern is analogous withchange of sound picture. This can be measuredby the accelerometer

Frequency distribution in 2 sec time windows before and after occlusion

Frequency footprintBefore occlusion

Frequency footprintAfter occlusion

Frequency Analysis. Spectrogram

LAD Occlusion

-Analysis of periodic signals-Sensitive to small changes-Ischemia changes pattern

Data Analysis: Power Spectrum Difference

Occlusion Reperfusion

• Short-time FFT– 512 points moving window

• Difference relative first time window– Calculated for each window

2512

0))(( ref

ntot PSnPSrSignalPowe −∑=Δ

=

Detection of 1 minute ischemiaLAD-region CX-region

Diff

eren

ceva

lue

Diff

eren

ceva

lue

Relative Energy in Fundamental Frequency

Occlusion 1: 600s Occlusion 2: 860s Occlusion 3: 926s

Conclusions

Accelerometer sensors are suitable for monitoring heart movements

Occlusion of the LAD artery induces immediate regional changes in heart movement patterns that are detected by this sensor

Status and future workWe have

– Prototype sensor, system for data acquisition– Analysis algorithms that detect dysfunctional heart motion

We work on– Improved algorithms to reliably identify dysfunctional heart real-

time – Improved sensor design/miniaturisation– Biocompatible packaging– Key questions related to specificity, sensitivity

Future work– Human studies during beating heart coronary surgery– Incorporation of a miniaturised accelerometer sensor into a

temporary pace-maker electrode

Main TasksSensor design

• 3 axes, smallPackaging: Encapsulation and cabling

• Demanding environmentSignal interpretation

• How to recognize dysfunctional heart motion

Miniature Sensor• Prototype is too big

– Useful experiments– Too big for final product

• Microsensor– Performance tuned to

application• Good enough, small enough

– Not general accelerometer

• What performance is needed?– Data from experiments

• Status– Two PhD students– Five designs in test production

2 mm

New Sensor DesignsTest designs in MultiMEMS (SensoNor, Horten,

Norway)– 5 variations– Published and proprietary designs

• Adapted to our specifications• 4 seismic masses. Piezo-resistive sensing• 3x6 mm chips

Final sensor specifications– Package size: 2x2x5 mm– 3 axes. Range ±4g. Resolution 10 mg

Where and how?– High aspect ratio process– Capacitive or piezo-resistive?

Packaging• Challenges

– Biocompatible• Implanted ~7 days

– Electrical security• Near heart

– Cable and connections• ~ 10 leads, Ø<2 mm

– Strong and reliable

• Our resources– NEWPACK-project– One PhD student

2 mm