Brain perfusion magnetic resonance imaging using pseudocontinuous arterial spin labeling in 314 dogs and cats

Abstract

Background: Arterial spin labeling (ASL) is a noninvasive brain perfusion magnetic resonance imaging (MRI) technique that has not been assessed in clinical veterinary medicine.

Hypothesis/objectives: To test the feasibility of ASL using a 1.5 Tesla scanner and provide recommendations for optimal quantification of cerebral blood flow (CBF) in dogs and cats.

Animals: Three hundred fourteen prospectively selected client-owned dogs and cats.

Methods: Each animal underwent brain MRI including morphological sequences and ≥1 ASL sequences using different sites of blood labeling and postlabeling delays (PLD). Calculated ASL success rates were compared. The CBF was quantified in animals that had morphologically normal brain MRI results and parameters of ASL optimization were investigated.

Results: Arterial spin labeling was easily implemented with an overall success rate of 95% in animals with normal brain MRI. Technical recommendations included (a) positioning of the imaging slab at the foramen magnum and (b) selected PLD of 1025 ms in cats and dogs <7 kg, 1525 ms in dogs 7 to 38 kg, and 2025 ms in dogs >38 kg. In 37 dogs, median optimal CBF in the cortex and thalamic nuclei were 114 and 95 mL/100 g/min, respectively. In 28 cats, median CBF in the cortex and thalamic nuclei were 113 and 114 mL/100 g/min, respectively.

Conclusions and clinical importance: Our survey of brain perfusion ASL-MRI demonstrated the feasibility of ASL at 1.5 Tesla, suggested technical recommendations and provided CBF values that should be helpful in the characterization of various brain diseases in dogs and cats.

Keywords: ASL; MRI; arterial spin labeling; brain; cat; dog; perfusion.

FIGURE 1

Good diagnostic quality brain ASL perfusion images (gASL pattern) and corresponding anatomical MR images in a dog. A 1.5‐year‐old male intact Portuguese Sheepdog diagnosed with idiopathic facial nerve paralysis. Top row shows arterial spin labeling (ASL) cerebral blood flow (CBF) color map transverse images. Highly perfused anatomical structures are depicted in red, moderately perfused structures are depicted in green, and poorly to not perfused structures are depicted in dark blue to black. Bottom row shows corresponding T2‐weighted transverse images obtained at the same anatomic levels as ASL images. A, Frontal lobes level: presylvian groove (blue arrow), cingulate gyrus (yellow arrow). B, Parietal lobes level: cingulate gyrus (yellow arrow) and caudate nucleus (orange arrow), and internal capsule (white asterisk). C, Thalamus level: middle suprasylvian groove (blue arrow), thalamic nucleus (orange arrow), and piriform lobe and parahippocampal gyrus (green arrow). D, Mesencephalon level: mesencephalic nucleus (orange arrow), marginal groove (blue arrow), and corona radiata (white asterisk). E, Cerebellum level: cerebellar vermis (blue arrows), cerebellar cortex (yellow arrow), lingula of cerebellum (green arrow), and cerebellar white matter (white asterisk)

FIGURE 2

Flow chart of the entire study and flow chart of phase 3 of the study. Group mid‐C2: animals enrolled during phase 1 (slab positioned over the mid‐length of the second cervical vertebra). Group foramen magnum (FM): animals enrolled during phases 2 and 3 (slab positioned over the foramen magnum). pASL‐U: arterial spin labeling (ASL) pattern indicating a unilateral lack of brain perfusion signal (poor diagnostic quality ASL study). gASL: ASL pattern indicating a good diagnostic quality ASL study. pASL‐B: ASL pattern indicating a bilateral lack of brain perfusion signal (poor diagnostic quality ASL study). Optimal postlabeling delay (PLD): PLD value (1025, 1525, or 2025 ms) that provided, for a given dog in phase 3, a gASL pattern and the highest value of cerebral blood flow measured in the prosencephalic cortex. Phase 1: slab at mid‐C2 and PLD set at 1025 ms; Phase 2: slab at FM and PLD set at 1025 ms; Phase 3: slab at FM and PLD set at three different values (1025, 1525, and 2025 ms). n, number of animals (dogs and cats)

FIGURE 3

Box plot showing weight distribution relative to optimal PLD in 37 dogs with normal brain MRI. Optimal postlabeling delay (PLD) was determined in dogs of phase 3 that had normal brain magnetic resonance imaging (MRI) and gASL pattern with at least 1 of the PLDs tested

FIGURE 4

Box plot showing age distribution relative to optimal PLD in 37 dogs with normal brain MRI. Optimal postlabeling delay (PLD) was determined in dogs of phase 3 that had a normal brain magnetic resonance imaging (MRI) and a gASL pattern with at least 1 of the PLDs tested

FIGURE 5

Box plot showing weight distribution relative to optimal PLD in 27 dogs with pathological brain MRI. Optimal postlabeling delay (PLD) was determined in dogs of phase 3 that had pathological brain magnetic resonance imaging (MRI) and gASL pattern with at least 1 of the PLDs tested

FIGURE 6

Box plot showing age distribution relative to optimal PLD in 27 dogs with pathological brain MRI. Optimal postlabeling delay (PLD) was determined in dogs of phase 3 that had pathological brain magnetic resonance imaging (MRI) and gASL pattern with at least 1 of the PLDs tested