• Aucun résultat trouvé

Development and Identification of a Closed-Loop Model of the Cardiovascular System Including the Atria

N/A
N/A
Protected

Academic year: 2021

Partager "Development and Identification of a Closed-Loop Model of the Cardiovascular System Including the Atria"

Copied!
35
0
0

Texte intégral

(1)

DEVELOPMENT AND IDENTIFICATION

OF A CLOSED-LOOP MODEL

OF THE CARDIOVASCULAR SYSTEM

INCLUDING THE ATRIA

A.  PIRONET*, J. A. REVIE**, S. PAEME*,

P. C. DAUBY*, J. G. CHASE**, T. DESAIVE*

* CARDIOVASCULAR RESEARCH CENTER, UNIVERSITY OF LIEGE, LIEGE, BELGIUM

** DEPARTMENT OF MECHANICAL ENGINEERING, UNIVERSITY OF CANTERBURY, CHRISTCHURCH, NEW ZEALAND

8TH IFAC SYMPOSIUM ON BIOLOGICAL AND MEDICAL SYSTEMS

(2)

•  Initial CVS model

•  Introduction of the atria

•  Parameter identification

•  Results

CONTENTS

1

(3)

Aorta& Right ventricle Pulmonary artery Pulmonary& vein& Vena cava Le1& ventricle&

Flow resistance

Valve

METHODS

2

ORIGINAL CVS MODEL

(4)

P

1

P

2

R

Q

P

1

– P

2

R

Q =

P

1

– P

2

R

Q =

if P

1

> P

2

0

otherwise

Valve:

P

1

P

2

R

Q

No valve:

METHODS

2

ORIGINAL CVS MODEL

1

(5)

P

=

E(t)

V

Other chambers:

Cardiac chambers:

Le1&

ventricle& ventricle&Right&

P

=

E

V

Pulmonary&

vein& Aorta&

Pulmonary&

artery& Vena&cava&

METHODS

2

ORIGINAL CVS MODEL

(6)

Continuity equation:

V

Q

out

Q

in

V = Q

.

in

- Q

out

METHODS

2

ORIGINAL CVS MODEL

1

(7)

Pulmonary artery Pulmonary

vein atrium Left

Right

ventricle atrium Right Vena cava Aorta Left

ventricle

Flow resistance

Valve

+ 1 Active chamber + 1 Resistance + 1 Active chamber + 1 Resistance

METHODS

2

INTRODUCTION OF THE ATRIA

(8)

Left atrial pressure

(mmHg)

METHODS

2

INTRODUCTION OF THE ATRIA

2

16

12

8

12

8

6

10

Left atrial volume (mmHg)

One heartbeat

Filling

a

v

(9)

Left atrium

P(V,

t)

=

P

a

e(t) + E

V

(1 – e(t))

Atrial

volume pressure Atrial

METHODS

2

INTRODUCTION OF THE ATRIA

(10)

Left atrium

P(V,

t)

=

P

a

e(t) + E

V

(1 – e(t))

Atrial

volume pressure Atrial Active component

e(t) 1

METHODS

2

INTRODUCTION OF THE ATRIA

(11)

Left atrium

P(V,

t)

=

P

a

e(t) + E

V

(1 – e(t))

Atrial

volume pressure Atrial Passive component

e(t) 0

METHODS

2

INTRODUCTION OF THE ATRIA

(12)

Left atrium

P(V,

t)

=

P

a

e(t) + E

V

(1 – e(t))

Atrial

volume pressure Atrial Driver function

e(t) in [0, 1]

METHODS

2

INTRODUCTION OF THE ATRIA

(13)

Left atrium

P(V,

t)

=

P

a

e(t) + E

V

(1 – e(t))

METHODS

2

INTRODUCTION OF THE ATRIA

2

Pa A-wave peak pressure Maximum diastolic ventricular pressure Offset (atrio-ventricular pressure gradient)

(14)

Left atrium

P(V,

t)

=

P

a

e(t)

+ E

V

(1 –

e(t)

)

Driver function e(t) in [0, 1]

METHODS

2

INTRODUCTION OF THE ATRIA

2

Time of atrial contraction Time of maximum diastolic ventricular pressure

(15)

Left atrium

P(V,

t)

=

P

a

e(t)

+

E

V

(1 –

e(t)

)

Driver function e(t) in [0, 1]

METHODS

2

INTRODUCTION OF THE ATRIA

2

Width of driver

function Width of driver function Fixed Atrial passive

(16)

Model simulation Reference measurement

XX

METHODS

2

PARAMETER IDENTIFICATION

3

(17)

Initial parameter value Model simulation Reference measurement New parameter value

XX

METHODS

2

PARAMETER IDENTIFICATION

3

(18)

Pulmonary artery

Left atrium

Right

ventricle atrium Right

Aorta Left ventricle Fixed pulmonary vein pressure Fixed vena cava pressure

P

vc

P

pu

METHODS

2

PARAMETER IDENTIFICATION

3

(19)

Identification of the

systemic model Identification of the pulmonary model

Identification of ventricular elastances Computation of ventricular interaction

METHODS

2

PARAMETER IDENTIFICATION: STEP 1

(20)

Dataset 1 Dataset 2 Dataset N

Average valve resistances

METHODS

2

PARAMETER IDENTIFICATION: STEP 2

(21)

Identification of the

systemic model Identification of the pulmonary model

Identification of ventricular elastances Computation of ventricular interaction Fixed valve resistances

METHODS

2

PARAMETER IDENTIFICATION: STEP 3

(22)

Pulmonary artery

Left atrium

Right

ventricle atrium Right

Aorta Left ventricle Fixed pulmonary vein pressure Fixed vena cava pressure

P

vc

P

pu

METHODS

2

PARAMETER IDENTIFICATION: STEP 4

(23)

Pulmonary artery Pulmonary vein Left atrium Right

ventricle atrium Right Vena cava Aorta Left

ventricle

METHODS

2

PARAMETER IDENTIFICATION: STEP 4

(24)

METHODS

2

•  Pig with induced pulmonary

embolism

•  Measurements: ventricular

pressures and volumes, aortic and

pulmonary artery pressure

•  Taken every 30 minutes, providing

10 data sets for identification

PORCINE EXPERIMENTS AND DATA

(25)

METHODS

2

•  Pig with induced pulmonary

embolism

•  Measurements: ventricular

pressures and volumes, aortic and

pulmonary artery pressure

•  Taken every 30 minutes, providing

10 data sets for identification

PORCINE EXPERIMENTS AND DATA

(26)

METHODS

2

•  Pig with induced pulmonary

embolism

•  Measurements: ventricular

pressures and volumes, aortic and

pulmonary artery pressure

•  Taken every 30 minutes, providing

10 data sets for identification

PORCINE EXPERIMENTS AND DATA

(27)

Time

0.47

0.97

RESULTS

3

PULMONARY RESISTANCE (PRU)

(28)

8.05

RESULTS

3

Time

3.67

3.35

1.11

Le1&

atrium& atrium&Right&

ATRIAL PRESSURES (mmHg)

(29)

•  Lumped models have well-known

limitations

•  Atrial model is load-dependent

•  Hypotheses formulated for the atria

have to be further validated

DISCUSSION

4

(30)

•  Lumped models have well-known

limitations

•  Atrial model is load-dependent

•  Hypotheses formulated for the atria

have to be further validated

DISCUSSION

4

(31)

•  Lumped models have well-known

limitations

•  Atrial model is load-dependent

•  Hypotheses formulated for the atria

have to be further validated

DISCUSSION

4

(32)

•  Atria are inserted in a model of the

CVS

•  Identification method is extended to

the atria

First step to real-time atrial

monitoring inside a full

cardiovascular system model

CONCLUSION

5

(33)

•  Atria are inserted in a model of the

CVS

•  Identification method is extended to

the atria

First step to real-time atrial

monitoring inside a full

cardiovascular system model

CONCLUSION

5

(34)

•  Atria are inserted in a model of the

CVS

•  Identification method is extended to

the atria

First step to real-time atrial

monitoring inside a full

cardiovascular system model

CONCLUSION

5

(35)

Thanks for your attention!

Questions?

Références

Documents relatifs

We proposed a meta-algorithm, called unified gap-based exploration (UGapE), that unifies the two settings of the best arm(s) identification problem in stochastic multi-armed

Face à l’émergence de nouvelles BLSE et vu l’ampleur et la complexité du problème, des études épidémiologiques seront nécessaires pour évaluer de façon précise la

2b reaction of end-group of P-1 with pendent group of P-2 in the presence of a condensing agent 2c reaction of end-group of P-1 with pendent group of P-2 in presence of a

Calculer la solubilité de ce dernier en moles par litre puis en grammes

An advantage of this approach is given by its versatility: given a complex image of high resolution, the level set segmentation can be resolved at pixel level; in exchange, the

Acramov._ Séléction préparation des jeunes footballeurs._ (Traduit par : A. _ ﻊﻴﺯﻭﺘﻟﺍﻭ ﺭﺸﻨﻠﻟ ﻕﻭﺭﺸﻟﺍ ﺭﺍﺩ.. ﻱﺭﻜﺫﻟﺍ ﺝﺫﻭﻤﻨﻟﺍ ﻲﻓ ﺔﻠﺜﻤﻤﻟﺍ ﺔﻴﻭﻁﻠﺴﻟﺍ ﺔﺴﺭﺎﻤﻤ ) ﺏﻷﺍ (

There exists a dendrite, not reduced to a point, whose branch points all have orders in S and with the following property: for every n ∈ S, every arc contains an element of order

age neuronal soma size exhibited different profiles of postnatal changes in distinct layers of different subdivisions of the monkey entorhinal cortex.. Based on these results,