Related Experiment Videos
Automatic quantification of ejection fraction from gated myocardial perfusion SPECT
G Germano1, H Kiat, P B Kavanagh
1Department of Medical Physics and Imaging, Cedars-Sinai Research Institute, Cedars-Sinai Medical Center, Los Angeles, California 90048, USA.
Summary
A new automatic algorithm accurately measures left ventricular ejection fraction (LVEF) from myocardial perfusion SPECT images. This rapid method provides clinically useful data comparable to traditional radionuclide measurements.
Area of Science:
- Nuclear Medicine
- Cardiovascular Imaging
- Medical Imaging Analysis
Background:
- Left ventricular ejection fraction (LVEF) is a critical indicator of cardiac function.
- Quantitative assessment of LVEF from myocardial perfusion SPECT images can be challenging.
Purpose of the Study:
- To develop and validate a fully automated algorithm for measuring LVEF from gated 99mTc-sestamibi myocardial perfusion SPECT images.
Main Methods:
- A 3D algorithm was employed to segment the left ventricle (LV) and delineate endocardial and epicardial surfaces.
- The algorithm automatically calculated LV volumes and derived global ejection fraction (EF) from end-diastolic and end-systolic volumes.
- The method was tested on 65 patients and validated against first-pass radionuclide ventriculography.
Main Results:
- The automated LV segmentation and contouring achieved 100% success rate (65/65 studies).
- High agreement was observed between LVEF derived from 8-interval gated SPECT and first-pass data (r = 0.909).
- Excellent agreement was found between LVEF from 16-interval and 8-interval gated SPECT (r = 0.988).
Conclusions:
- The developed automatic algorithm is rapid and provides LVEF measurements that highly agree with conventional radionuclide methods.
- This automated approach offers clinically valuable information to supplement myocardial perfusion studies.