A gamma-distribution convolution model of ^99mTc-MIBI thyroid time-activity curves

Area of Science:

• Nuclear Medicine
• Radiopharmaceutical Kinetics
• Biophysical Modeling

Background:

• Thyroid time-activity curve (TAC) analysis is crucial for understanding radiotracer behavior.
• Traditional models may not fully capture complex distribution and washout kinetics.
• The gamma distribution convolution (GDC) model offers a more comprehensive approach.

Purpose of the Study:

• To present and apply a novel convolution approach using a GDC TAC model for thyroid imaging.
• To evaluate the GDC model's accuracy and performance against simpler models.
• To assess the GDC model's utility in simulating thyroid activity.

Main Methods:

• Developed a GDC TAC model by convolving two gamma distribution functions.
• Fitted the GDC model to thyroid region of interest (ROI) TAC data from 99mTc-MIBI SPECT imaging (1 min/frame for 90 min).
• Validated GDC model accuracy using Monte Carlo simulations and compared it with washout-only models.

Main Results:

• Clinical GDC models showed low fit errors (0.998%-1.82%) with mean residence times (MRTs) ranging from 212 to 699 min.
• Simulations demonstrated high accuracy (0.7% coefficient of variation, 0.5% accuracy error) for the GDC model.
• GDC models significantly outperformed washout-only models in MRT estimation (R-squared 0.0183-0.9395 vs. GDC).

Conclusions:

• The GDC model demonstrates high accuracy and plausible use for simulating thyroid activity in 99mTc-MIBI imaging.
• The GDC model surpasses the performance of washout-only models within a 90-minute acquisition period.
• Preliminary findings suggest the GDC model's potential for evaluating other kinetic models under various conditions.