Automatic postprocessing for the assessment of quantitative human myocardial perfusion using MRI.

Area of Science:

• Cardiovascular Imaging
• Medical Physics
• Radiology

Background:

• Quantitative determination of myocardial perfusion is crucial for diagnosing cardiac conditions.
• Current postprocessing methods for myocardial perfusion imaging are time-consuming, limiting clinical application.

Purpose of the Study:

• To develop and evaluate an automatic postprocessing algorithm for quantitative myocardial perfusion imaging.
• To compare the accuracy and efficiency of automatic versus manual postprocessing.

Main Methods:

• A retrospective study involving 75 cardiac MRI examinations (57 rest, 18 stress).
• Automatic postprocessing included image registration and segmentation to derive regional signal intensity time courses and perfusion values.
• Comparison with manual evaluation, including interobserver variability assessment in 10 cases.

Main Results:

• Automatic postprocessing was successfully applied to all datasets, yielding results comparable to manual evaluation.
• Variability between automatic and manual evaluation was 0.34 ml/g/min, similar to interobserver variability (0.35 ml/g/min).
• Processing time per slice was reduced from 25 minutes (manual) to less than 1 minute (automatic).

Conclusions:

• Automatic postprocessing provides accurate quantitative myocardial perfusion values equivalent to manual methods.
• Significant reduction in processing time facilitates routine clinical use of quantitative magnetic resonance perfusion imaging.
• This automation has the potential to make quantitative MRI perfusion a standard clinical procedure.