The collimator, which is an important part that determines the quality of SPECT imaging, has continued to be used since Anger first developed a gamma camera in the 1950s. The most common collimator used in nuclear medicine is a parallel-hole collimator, which has developed into a myriad of holes aligned in parallel with each other with a thin lead plate (height number cm) as a partition wall. The shape of the holes, e.g., hexagonal, round, and square, is determined by the production method of the collimator.

Collimators are classified into two major groups: low energy (LE) and medium energy (ME). The thickness of the partition wall (septa) of collimators is designed so that the ratio of the gamma rays passing through it is less than 5 %. Therefore, ME collimators have a thicker partition wall than LE collimators [5] to prevent the penetration of higher-energy gamma rays. The increased septal thickness of ME collimators contributes to increase image resolution and decrease scatter in the images compared to LE collimators but also degrades count sensitivity and spatial resolution. Moreover, nuclear laboratories are not always equipped with ME collimators, but use LE collimators instead because they are suitable for technetium (^99mTc), which is used most commonly in clinical practice.

Actually, in addition to LE and ME collimators, various types of collimators are also available, depending on the purpose of the study, in order to achieve good balance among sensitivity, resolution, and appropriate energy extent: low-energy high resolution (LEHR), low-energy high sensitivity, low-energy general purpose (LEGP), medium-energy general purpose, high-energy general purpose, and high-energy pinhole collimators [17, 19]. With respect to ¹²³I-MIBG cardiac scintigraphy, most clinical and published studies have used LE collimators, such as LEGP or LEHR collimators.

Iodine-123 predominantly emits energy of 159 keV (97 %), which is an optimal imaging energy for an LE collimator, but also emits high-energy photons of more than 400 keV (approximately 2.9 %), with the main contributing photons at 529 keV. Higher-energy ¹²³I photons, mostly at 529 keV, can penetrate the thinner septal wall and contaminate 159 keV imaging data [5, 12, 20] (Fig. 14.1). Thus, an LE collimator, which has a thinner septal wall, can show a lower H/M ratio than an ME collimator. On the basis of these findings, recent guidelines on MIBG imaging from the European Council of Nuclear Cardiology recommend using an ME collimator, which has a thicker septal wall than an LE collimator, for MIBG imaging to minimize scattered radiation noise from high-energy emissions [5], but this recommendation is not usually followed in clinical practice. In addition, a lot of previous studies have used an LE collimator.

Clinically, it is a considerable problem that differences in the collimator can influence the H/M ratio profoundly. To compensate for the resulting corruption of image quantitation related to LE collimator penetration by higher-energy photons, Chen et al. developed a mathematical technique (iterative reconstruction with deconvolution of septal penetration) that appears to improve the quantitative accuracy of cardiac ¹²³I-MIBG uptake in reference to a phantom standard [21]. However, clinical applications have yet to be determined.

On the contrary, Nakajima et al. have endeavored to standardize the heart-to-mediastinum (H/M) ratio assessed by using planar images. They designed a simple and reproducible phantom standard for planar image acquisition [20, 21]. The phantom consisted of multiple slices of the heart, lung, mediastinum, and liver containing a radioisotope as a single component. Using these two phantoms, four H/M ratios (anterior and posterior views for each) were obtained to calculate the regression equation. A linear regression equation was analyzed using the formula y − 1 = Ki × (x − 1). The first step was to convert the H/M ratio generated by the LE collimator to the reference value based on mathematical theory using the conversion coefficient Ki. The second step was to convert this H/M ratio to a standardized H/M ratio with the conversion coefficient Ks, which was defined as the average K value for a typical ME collimator. They verified the initial standardization efforts in ten centers, suggesting that this phantom method could be applied to calibrate the results from ME collimators and LE collimators [22].

More recently, much larger Japanese multicenter studies including 84 institutions for the measurement of planar H/M ratios using standard nuclear cameras from a variety of vendors and collimators also confirmed and extended these findings [19]. On the basis of phantom studies, a conversion coefficient of 0.88 was determined to integrate H/M ratios from all acquisition conditions. Using the reference H/M ratio and conversion coefficients for the system can convert an H/M ratio under various conditions converted to the standard one, irrespective of the collimator used. They also demonstrated that two H/M ratios from one phantom could be comparable to one from two phantoms because the conversion coefficient showed a significantly high correlation with an R² of 0.997. Standardization of the H/M ratio for ¹²³I-MIBG among various scinticamera-collimator combinations will be useful for not only clinical practice but also multicenter studies.