oral presentation at stacom10

26
Atlas-based Quantication of Myocardial Motion Abnormalities: Added-value for the Understanding of CRT Outcome? STACOM-CESC Workshop, MICCAI 2010 Beijing– 20/09/2010

Upload: mathieu-de-craene

Post on 25-Jun-2015

695 views

Category:

Technology


2 download

DESCRIPTION

N. Duchateau, M. De Craene, G. Piella, C. Hoogendoorn, E. Silva, A. Doltra, L. Mont, M.A. Castel, J. Brugada, M. Sitges, and A.F. Frangi. Atlas-based Quantification of myocardial motion abnormalities: Added-value for the understanding of CRT outcome?. In Statistical Atlases and Computational Models of the Heart (STACOM). MICCAI Workshop., 2010. http://www.dtic.upf.edu/~mde/pdf/stacom10/DuchateauStacom10.pdf

TRANSCRIPT

Page 1: Oral presentation at STACOM10

Atlas-based Quantication of Myocardial MotionAbnormalities: Added-value for theUnderstanding of CRT Outcome?

STACOM-CESC Workshop, MICCAI 2010Beijing– 20/09/2010

Page 2: Oral presentation at STACOM10

1a

Atlas-based quantification of motion abnormalities

Healthy subjects

Radial velocity

Long. velocity

Atlas

variance

average

(mm/s)

(mm/s)

Page 3: Oral presentation at STACOM10

1b

Atlas-based quantification of motion abnormalities

Radial velocity

Long. velocity

Atlas

d = ???

Healthy subjects

Patient to study

p-value (log scale)

(mm/s)

(mm/s)

Page 4: Oral presentation at STACOM10

New quantitative indexes [quantification of motion abnormalities]

Statistical atlas Automatic Reproducible

Contributions

2

In this work: Mechanisms involved in CRT response Quantification before and after the therapy

Added-value for clinical studies Accurate, automatic Generic methods applicable to almost any

imaging modality studied parameter and mechanism

Page 5: Oral presentation at STACOM10

[1] Stellbrink et al. , EHJ Suppl. 2004 [5] Parsai et al., EHJ 2009 [2] Chung et al. , Circulation 2008 [6] De Boeck et al., EJHF 2009[3] Fornwalt et al. , JASE 2009 [7] Voigt et al., EHJ 2009[4] Voigt, EHJ 2009

Lack of reproducibility in large scale studies [1]

Is there a “universal” index? [2,3,4]

Changing the strategy?

Patient classification into specific etiologies of HF [5]

Correction of specific mechanisms of dyssynchrony conditions response

Predicitive value of specific classes

• Septal flash [5]

• Septal rebound stretch [6]

• Apical transverse motion [7]

3

Why quantifying abnormalities? CRT context

Need to accurately characterize these patterns

Page 6: Oral presentation at STACOM10

Fig.3: Septal flash mechanismWhat is a “septal flash” ?

Healthy volunteer CRT candidate with SF

4

Parsai, Bijnens et al., EHJ 2009

Page 7: Oral presentation at STACOM10

3

Effect of CRT on septal flash

Pre-CRT Follow-up (6 months)

5

Page 8: Oral presentation at STACOM10

3

Effect of CRT on septal flash

Pre-CRT Follow-up (6 months)

Pre-CRT

5

Page 9: Oral presentation at STACOM10

3

Effect of CRT on septal flash

Follow-up(6 months)

Pre-CRT Follow-up (6 months)

Pre-CRT

5

Page 10: Oral presentation at STACOM10

Plan

Atlas pipeline

Relevance of the atlas population

Clinical outcome after CRT

Page 11: Oral presentation at STACOM10

Healthy subjects

Atlas of “normality”

Registration-based tracking

Spatio-temporal

normalization

Group statistics: average, covariance, …

Myocardial velocities

8a

[9] Duchateau et al. , MICCAI 2009

Construction of an atlas of “normality”[9]

Page 12: Oral presentation at STACOM10

Healthy subjects

Atlas of “normality”

Registration-based tracking

Group statistics: average, covariance, …

8b

[9] Duchateau et al. , MICCAI 2009

Construction of an atlas of “normality”[9]

Normalized timescaleInital ECG

Temporal normalization

Spatio-temporal

normalization

Page 13: Oral presentation at STACOM10

Healthy subjects

Atlas of “normality”

Registration-based tracking

Group statistics: average, covariance, …

8c

[9] Duchateau et al. , MICCAI 2009

Construction of an atlas of “normality”[9]

Spatio-temporal

normalization

Spatial reorientation

Page 14: Oral presentation at STACOM10

Atlas of “normality”

Population of CRT

candidates

9

Statistical distance to “normality”

Healthy subjects

d = ???

Statistical distance = p-value associated to Mahalanobis distance

LOW p-value = HIGH abnormality

Page 15: Oral presentation at STACOM10

Plan

Atlas pipeline

Relevance of the atlas population

Clinical outcome after CRT

Page 16: Oral presentation at STACOM10

2D echo, 4-chamber view

11

Data available

21 Healthy volunteers

60 frames/s0.24 x 0.24 mm2

88 candidates OFF / ON / FU (11+/- 2 months)EF < 35%, QRS duration > 120ms, and (or) NYHA class III-IV

60 frames/s0.24 x 0.24 mm2

CRT response:Clinical 6min walking test increase ≥ 10%

Echocardiographic LV end-systolic volume reduction ≥ 15% or NYHA class reduction ≥ 1 point

Page 17: Oral presentation at STACOM10

13

Is the atlas representative of “normality”?

• Non-dilated hearts• No antecedent of cardiac dysfunction• Normal baseline characteristics• Young (30 +/- 5)

How many subjects?

Page 18: Oral presentation at STACOM10

14

Is the atlas representative of “normality”?

• Non-dilated hearts• No antecedent of cardiac dysfunction• Normal baseline characteristics• Young (30±5)

How many subjects?

Statistical distribution assumption

d = ???

Statistical distance = p-value associated to Mahalanobis distance

Gaussianity tests:Shapiro-Wilk (SW) and Lilliefors (LF)

Page 19: Oral presentation at STACOM10

Plan

Atlas pipeline

Relevance of the atlas population

Clinical outcome after CRT

Page 20: Oral presentation at STACOM10

Inward Outward

10a

Data representation

Temporal evolution at a fixed anatomical point

p-value (log scale)

Local maps at fixed time t

p-value (log scale)

Red = large abnormality

Page 21: Oral presentation at STACOM10

Data representation

Spatiotemporal maps of abnormality

Blue = Inward (vp<0)Red = Outward (vp>0)

Base

ApexTime

IVC Systole Diastole

Inward Outward

10b

p-value (log scale)

Page 22: Oral presentation at STACOM10

Spatiotemporal quantification of abnormalities

CRT #9Septal flash

CRT #8Septal flash

CRT #12Left-right

interaction

15a

Blue = Inward (vp<0)Red = Outward (vp>0)

Local p-value * sign of radial velocity

(log scale)

???

IVC Systole Diastole

OFF

Follow-up

Page 23: Oral presentation at STACOM10

Spatiotemporal quantification of abnormalities

CRT #9Septal flash

CRT #8Septal flash

CRT #12Left-right

interaction

15b

Blue = Inward (vp<0)Red = Outward (vp>0)

IVC Systole Diastole

OFF Follow-upLocal p-value * sign of

radial velocity(log scale)

Page 24: Oral presentation at STACOM10

Reduction of specific abnormalities (SF)

17

Correction of SF = High predictive value

p-value(log scale)

Page 25: Oral presentation at STACOM10

Conclusions

18

Added-value for clinical studies Accurate, automatic Information still available at every location (x,t) [not heart segments only] Generic methods applicable to almost any

imaging modality studied parameter and mechanism

Clinical conclusionsSimilar observations than in previous clinical studies [5,6]

Observation of global abnormalities leads to limited conclusions

Correction of specific abnormalities (e.g. SF) = high predictor of response

Further work = extension to strain measurements (influence of local infarction)

[5] Parsai et al., EHJ 2009[6] Parsai et al., EHJ 2009

Page 26: Oral presentation at STACOM10

CISTIB, Universitat Pompeu Fabra Image registration team M. De Craene, G. Piella

Hospital Clínic, Barcelona E. Silva, A. Doltra, M. Sitges, B. H. Bijnens

Acknowledgements

Related worksAtlas construction: N. Duchateau, M. De Craene, E. Silva, M. Sitges, B. H. Bijnens, and A. F. Frangi “Septal Flash Assessment on CRT Candidates based on Statistical Atlases of Motion” MICCAI’09 LNCS 5762 (pp.759-766)

Quantification of CRT outcome: N. Duchateau, A. Doltra, E. Silva, M. De Craene, G. Piella, L. Mont, Ma A. Castel, J. Brugada, M. Sitges, and A. F. Frangi “Added value of a statistical atlas-based quantification of motion abnormalities for the prediction of CRT response” EuroEcho 2010 Randers – Lecture rooms – 09/12/2010

19