Improved dose-volume histogram estimates for radiopharmaceutical therapy by optimizing quantitative SPECT...

Area of Science:

• Medical Physics
• Nuclear Medicine
• Radiotherapy Physics

Background:

• Accurate dose distribution is vital for effective radiopharmaceutical therapy.
• Three-dimensional (3D) dosimetry and dose-volume histograms (DVHs) are essential for treatment planning.
• Quantitative SPECT (QSPECT) image reconstruction is critical for reliable 3D dosimetry but is affected by noise and partial volume effects (PVEs).

Purpose of the Study:

• To systematically investigate the impact of OS-EM based QSPECT image reconstruction on DVH estimates.
• To evaluate the influence of noise, PVEs, and image artifacts on DVH accuracy.
• To determine optimal reconstruction and regularization parameters for accurate DVH estimation in radiopharmaceutical therapy.

Main Methods:

• Utilized a modified 3D NURBS-based Cardiac-Torso (NCAT) phantom with realistic organ activity and biokinetics.
• Generated projections using Monte Carlo (MC) simulations with 50 noise realizations.
• Reconstructed activity images using QSPECT with compensation for attenuation, scatter, and collimator-detector response (CDR), then estimated dose rate distributions and calculated cumulative DVHs.

Main Results:

• Noise, PVEs, and CDR compensation artifacts degraded DVH estimates.
• Low-pass filtering and early iterative termination reduced noise/artifact impact but increased PVE degradation.
• Optimal reconstruction parameters varied by organ size and time post-administration, with single parameter sets yielding near-optimal results.

Conclusions:

• Optimal reconstruction and regularization parameters are crucial for accurate DVH estimation in QSPECT.
• Reconstruction parameters must be tailored to specific organs and time points for best results.
• While single parameter sets can provide near-optimal DVHs, individualized optimization enhances accuracy in radiopharmaceutical therapy dosimetry.