A new, semi-automatic, platform-independent method to semiquantify 123I-FP-CIT uptake: inter-operator variability in 86 cases

Aim: semi-quantification of 123I-FP-CIT uptake is a challenging issue but it is nowadays mandatory to allow comparison with normal reference values, assessment of severity, progression of disease and monitor therapeutic effects. Aim of the study was to realize a JAVA-based program (SUQS: Striatal Uptake Quantification System), therefore platform-independent (tested on Win, Mac an Linux), to assess semiquantitatively the striatal uptake of 123I-FP-CIT. Striatal uptake have been also converted in percent of injected dose. Inter-operator variability of the proposed method has been tested in 86 patients. Material and Methods: Charts of 86 consecutive patients (43 males, 43 females, mean age 69.4±9 years, range 42-88) who underwent 123I-FP-CIT scintigraphy from 2006 to 2007 were reconstructed using 20 maximum likelihood expectation maximization iterations. SPET was performed using a triple-head gamma-camera equipped with a low-energy, ultra-high-resolution collimator (step and shoot mode, 30 sec/step, 120 degrees, 128x128 matrix size, mean rotational radius 13.9±1.1). The same 3D Butterworth post-filter was then applied to each study. Attenuation correction was not applied to avoid bias on quantification. Transaxial tomographic slices (3.5 mm thick), exported as DICOM files, have been loaded in SUQS by two blinded, experienced nuclear medicine physicians. Program automatically generates striatal contour on all transaxial slices and obtains the total counts in the striatum (corrected for background and splitted into right and left striatum, caudatum and putamen). Operator can vary the sensitivity of the system setting a threshold when visual feedback suggests it. A fully tissue-equivalent anthropomorphic phantom (RSD) has been used to measure conversion coefficent (from counts to MBq) at rotational radius varying from 13 to 20. Therefore SUQS, using counts and radius, returns striatal uptake as percent of injected dose (corrected for residual activity and decay). All obtained data (net counts, total pixel per area, percents ecc) are exportable as comma separated values. Results: Proposed method was feasible in all 86 patients and reliable also at very low signal to background ratio. A mean time of about 2.5 minutes was required to elaborate each study. Measured uptakes, expressed as percent of injected dose, were very similar for both operators (right striatum:OP1=0.01% to 0.44%; OP2=0.01% to 0.52%, left striatum:OP1&OP2=0.01% to 0.5%). A Bland&Altman plot revealed good inter-observator reproducibility: mean of differences was 0.04 MBq and coefficient of reproducibility 0.07 MBq. Conclusions: We proposed a new method of striatal semi-quantification. SUQS is a fast, feasible, reliable, semi-automatic, platform-independent method with good inter-operator reproducibility.