Semiautomatic Quantification of 99mTc-TRODAT-1 SPECT Images in Patients With Idiopathic Parkinson's Disease

Abstract

Background and Purpose

^99mTc-TRODAT-1 SPECT imaging is an imaging technique, more commonly used in Asia, to diagnose Parkinson's disease (PD). This study evaluates the use of automated three-dimensional volume-of-interest (VOI) analysis in diagnosing PD.

Methods

^99mTc-TRODAT-1 SPECT images of 76 patients (50 with PD and 26 without PD) were retrospectively analyzed. The specific binding ratio (SBR) was calculated using an automated program. Multiple linear regression and receiver operating characteristic curve analyses were performed to identify the factors that affect SBR value and determine the optimal cutoff values.

Results

Multiple regression analysis revealed disease status as the strongest predictor of SBR values, followed by age and sex. Receiver operating characteristic curve analysis demonstrated good diagnostic performance for the striatum (area under the curve [AUC] = 0.922), putamen (AUC = 0.922), and caudate (AUC = 0.881). Optimal cutoff values were determined for the striatum (0.515; sensitivity 88.5%, specificity 90.0%), putamen (0.445; sensitivity 92.3%, specificity 86.0%), and caudate (0.655; sensitivity 84.6%, specificity 90.0%).

Conclusions

Semiautomatic quantitative analysis of ^99mTc-TRODAT-1 SPECT using automated three-dimensional VOI shows excellent diagnostic performance in differentiating PD from non-Parkinson's cases. Age and sex significantly influence SBR values, suggesting the need for demographic-adjusted cutoff values in clinical practice.