Related Experiment Videos
[Automatic calculation of left ventricular volume and ejection fraction from gated myocardial perfusion SPECT--basic
Y Kinoshita1, I Nanbu, J Tohyama
1Department of Radiology, Nagoya Daini Red Cross Hospital.
Summary
This study validates a Quantitative Gated SPECT program for accurate left ventricular (LV) volume and ejection fraction calculation. The automated method shows good reproducibility and accuracy, though careful evaluation is needed for anterior or inferoposterior wall defects.
Area of Science:
- Nuclear Medicine
- Cardiovascular Imaging
- Medical Physics
Background:
- Quantitative Gated SPECT (QGS) is crucial for assessing left ventricular (LV) function.
- Accurate LV volume and ejection fraction (EF) calculation relies on precise cardiac surface contour tracing.
- Automated methods aim to improve efficiency and reproducibility in SPECT analysis.
Purpose of the Study:
- To evaluate the accuracy of an automated Quantitative Gated SPECT program for calculating LV volume and EF.
- To assess the impact of different projection types (180-degree L-type vs. 360-degree opposed) on accuracy.
- To determine the reliability of automated contour tracing in the presence of cardiac defects.
Main Methods:
- Utilized cardiac phantoms filled with 99mTc-solution for accuracy assessment.
- Acquired SPECT data using 180-degree L-type and 360-degree opposed projections.
- Employed an automated program for LV contour tracing, volume, and EF calculation.
- Investigated the effect of varying prefilter cut-off values on accuracy.
Main Results:
- The automated QGS program calculated LV volume and EF within 3-4 minutes with sufficient reproducibility.
- An optimal prefilter cut-off value of 0.45 yielded LV volumes of 93% (L-type) and 95.9% (opposed-type) of actual volumes.
- LV volume was significantly smaller in L-type acquisition compared to opposed-type (p < 0.05).
- Automated tracing showed fair accuracy for septal and lateral wall defects but deviated anteriorly for anterior and inferoposterior wall defects, affecting LV volume estimation.
Conclusions:
- The automated QGS program provides a reproducible and reasonably accurate method for LV volume and EF assessment.
- Acquisition type influences LV volume measurements, with opposed-type projections showing slightly higher accuracy.
- Careful interpretation is necessary for patients with anterior or inferoposterior wall defects due to potential tracing deviations.