Purpose
To evaluate the effect of introducing a CNR-optimized composite image on inter-observer variability of myocardial contour determination across four observers compared to subjectively-chosen image.
Short introduction:
Myocardial T2* measured by bright-blood MRI has been shown to be correlated to iron concentration [1].
Manual myocardial contouring contributes variability to the determination of cardiac iron by T2* MRI.
A single short axis image,
corresponding with a single echo time (TE) from multi gradient echo (MGE) short axis MRI images,
is generally used to draw left ventricle (LV)...
Methods and Materials
1.
Patient information:
n = 36 SA slices from bright-blood T2* scans of 21patients (9 haematology [median: 30.11 ms,
range: 9.44-40.71 ms],
12 suspected cardiomyopathy [median: 33.04 ms,
range: 24.18-43.33 ms]).
2.
Highlighted MGE MRI sequence in the study:
Single breath hold,
1.5 T,
8 TE (2.59-18.20 ms,
2.23 ms increment).
3.
Observers who drew myocardium contour:
2 observers with >2 years experience
2 observers with < 1 year experience but with profound knowledge of short axis myocardium anatomy
4.
SA image selection to segment...
Results
Regardless of subjective selection,
inter-observers correspondence in CNR-optimized composite image shows better reproducibility (DSC improvement) of short axis LV-myocardial contour segmentation (Table 1).
For observers selecting the same image,
there is no significant difference between experienced observers with regards to the goodness of DSC with the composite image method but the correspondence between the composite image ROIs shows DSC improvement for less experienced observers (Table 1).
However,
when the selected images differ,
the correspondence between the composite image ROIs significantly exceeds that between the subjectively...
Conclusion
CNR-optimized composite image increases inter-observer consistency of contour detection and reproducibility of myocardial iron T2* calculation.
References
Nilesh R,
et al.
J of Magn Reson Imaging 2006,
23:9–16
Fayed N,
et al.
Acad Radiol 2006,
13:728–737
Zou KH, et al.
Acad Radiol 2004,
11(2): 178–189.
Personal Information
Corresponding author:
Pandji Triadyaksa
University of Groningen,
University Medical Center Groningen,Center for Medical Imaging-North East Netherlands,
Department ofRadiology.
Email:
[email protected] ;
[email protected]
Funding : Directorate General of Higher Education (DGHE) of Indonesia