Optimum detector arrangements for in-beam PET with direct time-of-flight

Due to its effectiveness in monitoring highly tumourconformed charged hadron therapy, in-beam positron emission tomography (in-beam PET [1]) is strongly desirable at charged hadron facilities under construction [2] or planning [3]. We present, therefore, optimized arrangements of the γ-ray detectors for next-generation in-beam PET scanners. A versatile, fully 3D, maximum likelihood expectation maximization (MLEM) algorithm coupled to a simulation routine was applied to several high-resolution closed-ring or dual-head tomographs. In order to evaluate the quality of images obtained with several camera configurations in real therapeutic situations, β-activity distributions calculated from a treatment plan with the PosGen Monte-Carlo code [4] were considered. Results regarding the optimisation of in-beam PET for monitoring head-andneck irradiation can be found in [5, 6, 7]. Here we refrain to the more challenging situation, due to the large image volume, of monitoring the irradiation of the pelvis. Fig. 1 depicts the positioning strategy for monitoring irradiation of the pelvis with in-beam PET and Fig. 2 shows the corresponding reconstructed images obtained without (top and middle rows) and with (bottom) time-of-flight (TOF) capable detectors. The sagittal views in Fig. 2 show the rectum of the patient, a radiosensitive organ, lying adjacent to the irradiated tumour and, therefore, elucidate how in-beam PET monitoring brings important information to the radiooncologist. Clearly, the closed-ring detector configuration (top row) yields the best images and its implementation feasibility is discussed in [5, 6]. Although the field of view of the dual-head tomograph with φ = 46 ◦ (middle row) covers the irradiated volume completely, the corresponding reconstructed images contain less information due to the image degradation arising from the gaps between the two detector heads [5, 6]. This degradation is refrained [5, 7] if TOF-capable detectors with sufficiently good coincidence time resolution are used (bottom row). In summary, in-beam PET monitoring of large fields benefits most from a closed-ring tomograph if ultra-fast TOF-capable detectors are not available at large production scales. A non-TOF dual-head tomograph with φ = 46 yields satisfactory results for monitoring head-and-neck irradiation due to the smaller target volume (not shown). References [1] W. Enghardt et al., “Charged hadron tumour therapy monitoring by means of PET”, NIM A 525 (2004) 284. [2] H. Eickhoff et al., “HICAT The German hospital-based light ion cancer therapy project”, PAC’03, Portland, OR, p. 694.