Cardiac MRI and Myocardial Infarction: the role of ......Cardiac MRI and Myocardial Infarction: the role of Mathematical Modelling Kenneth Mangion, Clinical Research Fellow Developments

Cardiac MRI and Myocardial Infarction: the role of Mathematical Modelling

Kenneth Mangion,

Clinical Research Fellow

Developments in Healthcare Imaging

Cambridge, 19/04/17

Body text

Acute

STEMI Angiography Primary PCI Stent

Velagaleti et al., Circulation 2008

Coronary

Occlusion

Body text

Target

identifiable

difference

Sample size

reduction by

CMR

EDV (ml/m2) 10 74%

ESV (ml/m2) 10 83%

SV (ml/m2) 10 84%

LVEF (%) 3 87%

Mass (g/m2) 10 90%

Grothues et al., 2004

Background: imaging biomarkers

Why use CMR?

Body text

Golden Jubilee National Hospital

BHF GCRC Maths & Stats

Body text

Cine

LVEF is an

indirect

measure of

contractility

Haemorrhage

occurs in a

subset

Infarct size

overestimated

MVO varies

dynamically

T2* mapping LGE

Occluded Circumflex

artery

Angio

Background: imaging biomarkers

Derived from multi parametric CMR

Dall’ Armelina et al., Circ. Cardiovasc. Imaging , 2011. Carrick et al., JACC Imaging, 2015

Requires eGFR

>30ml/min

14% STEMI

23% NSTEMI

have eGFR

A

B

LVEF

38%

LVEF

37.5%

LVEF

54.6%

LVEF

29.3%

6 months

post MI

Day 2

post MI

BHF- MRMI

STUDY

Infarct

36% LV

mass

Infarct

35% LV

mass

Background: imaging biomarkers

Case example

SofTMech

EPSRC Centre for Multiscale Soft Tissue Mechanics

Mathematical modelling

6 healthy volunteers 6 STEMI patients

Novel Biomarkers

Age: 52±9

EF: 66±3%

Age: 55±12

EF: 40±3%

Padmanabhan et al., 2010. Delles et al., 2010.

Need for personalised modelling

Case example

• Passive stiffness

• Contractility (Tref)

• Normalised Active

Tension (AT/SBP)

Healthy volunteer

STEMI patient

Need for personalised modelling

Image derived LV model

Healthy volunteer STEMI patient

Need for personalised modelling

4D LV models

Clinical data LV reconstruction

Strain estimation

Mesh generation

FEM simulation

Personalised

Parameters

Parameter

identification

optimisation

Model Personalisation

1 1.02 1.04 1.06 1.08 1.1 1.12 1.14 1.16 1.18 1.20

5

10

15

20

25

30

35

40

45

50stress in fiber orientation

HV1 HV2 HV3

HV4

HV5

HV6

MI1

MI2

MI3

MI4

MI5

MI6

Amount of fiber stretch

Cauch

y S

tress

(kP

a)

Amount of fiber stretch

Wall

str

ess (

KP

a)

STEMI patients

Healthy volunteers

Results (1)

Passive stiffness

Healthy volunteers 64.8±8kPa

STEMI patients 69±7kPa

Ave

rag

e A

ctive

Te

nsio

n

(KP

a)

p=0.15

Results (2)

Active Tension

Healthy volunteers 0.44 ± 0.05

p<0.01

STEMI patients 0.57 ± 0.05

AT

/SB

P (

KP

a/m

mH

g)

Results (3)

Normalised Active Tension

• 1. Development of novel biomarkers

• 2. Validation against more

established markers (LVEF, infarct

size)

• 3. Clinically accessible!

Cardiac MRI and Myocardial Infarction

The Role of Mathematical Modelling

Thank you for your attention

http://www.glasgowheart.org

http://www.softmech.org