Standard phantoms, which consisted of the heart, mediastinum, lung, and liver, were generated. Three correction methods were compared: TEW and two IDW methods (IDW0 and IDW1). Low-energy high-resolution (LEHR), medium-energy (ME), and 123I-specific low-medium-energy high-resolution (LMEHR) collimators were used. Clinical studies were performed in 10 patients. In the phantom study, the H/M ratio was significantly underestimated without correction, with both the LEHR and ME collimators (70 % and 88 % of the true value). When H/M with the LEHR collimator was divided by uncorrected H/M with the ME collimator, the ratio (mean ± SD) was 80 % ± 5 % , 98 % ± 5 % , 104 % ± 7 % , and 98 % ± 5 % for the no-correction, TEW, IDW0, and IDW1 methods, respectively. Clinical studies with the LEHR collimator after TEW and IDW correction (uncorrected average H/M ratio, 1.86 ± 0.23; TEW, 2.47 ± 0.46, P = .0015; IDW, 2.46 ± 0.46, P = .0017) provided comparable values to the uncorrected ME collimator (2.56 ± 0.46, P = NS vs TEW and IDW).
The H/M ratio with the ME collimator, after application of the TEW or IDW methods, was close to the theoretical value in the phantom study. However, the corrected H/M ratios with the LEHR collimator provided comparable H/M ratios to the uncorrected ME data in phantom and clinical studies.