Small nuclei identification with a hemispherical brain PET

Abstract

Background: To confirm the performance of the first hemispherical positron emission tomography (PET) for the brain (Vrain) that we developed to visualise the small nuclei in the deep brain area, we compared ¹⁸F-fluorodeoxyglucose (FDG) brain images with whole-body PET images.

Methods: Ten healthy male volunteers (aged 22-45 years) underwent a representative clinical whole-body PET, followed by Vrain each for 10 min. These two scans were initiated 30 min and 45 min after FDG injection (4.1 ± 0.5 MBq/kg), respectively. First, we visually identified the small nuclei and then compared their standardised uptake values (SUVs) with the participants' age. Next, the SUVs of each brain region, which were determined by applying a volume-of-interest template for anatomically normalised PET images, were compared between the brain images with the Vrain and those with the whole-body PET images.

Results: Small nuclei, such as the inferior colliculus, red nucleus, and substantia nigra, were more clearly visualised in Vrain than in whole-body PET. The anterior nucleus and dorsomedial nucleus in the thalamus and raphe nucleus in the brainstem were identified in Vrain but not in whole-body PET. The SUVs of the inferior colliculus and dentate gyrus in the cerebellum positively correlated with age (Spearman's correlation coefficient r = 0.811, p = 0.004; r = 0.738, p = 0.015, respectively). The SUVs of Vrain were slightly higher in the mesial temporal and medial parietal lobes than those in whole-body PET.

Conclusions: This was the first time that the raphe nuclei, anterior nuclei, and dorsomedial nuclei were successfully visualised using the first hemispherical brain PET. TRIAL REGISTRATION : Japan Registry of Clinical Trials, jRCTs032210086, Registered 13 May 2021, https://jrct.niph.go.jp/latest-detail/jRCTs032210086 .

Keywords: Brain PET; FDG; Healthy volunteer; Raphe nucleus; Thalamus.

Fig. 1

Photographs of whole-body PET (Discovery MI) (a) and hemispherical brain PET (Vrain) with a schematic view of the detector arrangement in the expanded view (b)

Fig. 2

Representative images of FDG of the brain using Vrain (a) and whole-body PET (b). Axial images from the left are at the level of the inferior colliculus (arrows), basal ganglia, centrum semi-ovale, and fronto-parietal area. All PET images were co-registered to T1WI MRI (c). The SUV is indicated by the colour scale

Fig. 3

FDG-PET images using Vrain are shown at the top (a–c) and those using the whole-body PET are shown in the middle (d–f). Images g–i show the T1WI MRIs. The centre of each crossbar is placed on the anterior nucleus (a, d, g) or the dorsomedial nucleus (b, e, h) in the thalamus. These are clearly visualised using Vrain (a, b) but not using the whole-body PET (d, e). The substantia nigra (arrows) and red nucleus (dashed arrows) are visualised using Vrain (c) with an expanded scale for the squared area on the right, but each region is not visualised separately using the whole-body PET (f). The SUV is indicated by the colour scale

Fig. 4

Sagittal views of the brainstem using Vrain (a), whole-body PET (b), and MRI (c). The right expanded images show the axial images at the level (dashed line in the sagittal image), which is 4 mm above the isthmus (red arrow). The midbrain part of the raphe nucleus (bold arrows) is depicted ventral to the inferior colliculus (dashed arrows) using Vrain (a), and the centre of the crossbar is placed at the right component of the raphe nucleus in the midbrain. The pontine part of the raphe nucleus is also visualised on sagittal view (black arrow) (a) at 2 mm below the isthmus (red arrow). The raphe nucleus is not visualised on the whole-body PET (b). The SUV is indicated by the colour scale

Fig. 5

Representative images of 4-mm-diameter volume-of-interest (VOI) manually placed on each structure. The plots of the averaged standardised uptake values (SUV) of these right- and the left-side VOIs and ages are presented with the images