Bridging physics and practice: evaluating sensitivity, septal penetration, and detector dead time in terbium-161 gamma-camera imaging.

Introduction/aim: Terbium-161 (Tb) has emerged as a promising therapeutic radionuclide, yet standardized imaging guidelines are lacking. This study aimed to characterize a SPECT/CT system, currently used in an ongoing clinical trial (BETA PLUS; NCT05359146), focusing on sensitivity, septal penetration, and dead-time effects.

Methods: Measurements were conducted on a Siemens Symbia Intevo system using two collimators: low-energy high-resolution (LEHR) and medium-energy low-penetration (MELP). Two energy windows were evaluated: 75 keV ± 10% and 48 keV ± 20%. Planar sensitivity and penetration were assessed using a Tb-filled Petri dish. Penetration fractions were determined as a function of distance for each collimator-window combination. Dead time was measured intrinsically for each detector using a set of Tb point sources. SPECT measurements of a homogenous cylinder phantom were performed to assess count rate performance and predict activity levels at which dead-time effects could occur. To evaluate the potential impact of dead time in patient imaging, SPECT projection data from patients treated with 1 GBq of [Tb]Tb-DOTA-LM3 (n = 8) was analyzed.

Results: Sensitivity was comparable for both collimators at 75 keV (LEHR: 15.7 cps/MBq, MELP: 18.5 cps/MBq) and increased at 48 keV (LEHR: 44.4 cps/MBq, MELP: 67.9 cps/MBq). Maximum penetration occurred at 75 keV with the LEHR collimator (7.5% at 10 cm). In acquired spectra, more than half of the detected counts (51.6%) appeared above the 75 keV window with LEHR, compared to only 12.2% with MELP. Dead-time analyses revealed non-linear detector responses at wide-spectrum count rates exceeding 93 kcps, corresponding to in-field activities of 1.4-2.0 GBq for LEHR and 1.7-2.2 GBq for MELP. The dead-time constant was determined to 0.42 µs for both detector heads, however, the maximum recorded count rate differed significantly (384 kcps vs. 546 kcps). The median and maximum wide-spectrum count rate for patients treated with [Tb]Tb-DOTA-LM3 was estimated to ~ 20 and ~ 40 kcps per GBq 3 h p.i., respectively, when imaged with LEHR, corresponding to a maximum estimated dead-time loss of 1.7%.

Conclusions: While high-quality Tb SPECT imaging is feasible, careful consideration is essential; the wide range of photons emitted will produce a higher wide-spectrum count rate as compared to Lu. The use of low-energy collimators increases penetration and scatter, impairing quantitative accuracy and elevating the wide-spectrum count rate, which may intensify dead-time effects. At therapeutic activity levels (e.g., 7.4 GBq), dead time should be closely monitored to ensure reliable quantification.