SPECT reconstruction using preprocessing masking for extra-cardiac uptake versus standard processing in...

Area of Science:

• Nuclear Medicine
• Medical Imaging
• Cardiology

Background:

• Extra-cardiac accumulation in 99mTc myocardial perfusion SPECT can cause artifacts overlapping the inferior wall, hindering accurate image interpretation.
• These artifacts arise from organs like the liver and gastrointestinal tract, obscuring crucial details of the myocardium.

Purpose of the Study:

• To quantitatively evaluate the effectiveness of a novel image reconstruction method, the masking process on unsmoothed images (MUS) method.
• To assess the MUS method's ability to reduce artifacts caused by extra-cardiac accumulation in 99mTc myocardial perfusion SPECT.

Main Methods:

• Retrospective analysis of 200 patients (400 scans) using 99mTc-sestamibi (MIBI) with a one-day stress-rest protocol.
• Image reconstruction using filtered back projection (FBP) and ordered subset expectation maximization with resolution recovery (OS-EM-RR), with and without the MUS method.
• Phantom studies and quantitative comparison of clinical data to assess separation capability and contrast improvement.

Main Results:

• The MUS method reduced visually unseparated cases from 15.5% to 3.5% compared to conventional methods.
• Phantom studies showed that larger and closer extra-cardiac accumulation degraded inferior wall contrast, with significant reduction when a 2 cm distance was maintained.
• Clinical images demonstrated significant improvement in relative contrast for the inferolateral/inferior wall (mid-ventricular) and inferoseptal/inferior wall (basal) with the MUS method.

Conclusions:

• The MUS method effectively separates extra-cardiac accumulation from the myocardium, improving contrast in the inferior wall.
• This novel method shows significant clinical utility, especially when extra-cardiac accumulation overlaps or is in close contact with the myocardium.
• The MUS method enhances diagnostic accuracy in 99mTc myocardial perfusion SPECT by mitigating image artifacts.