Reference cluster normalization improves detection of frontotemporal lobar degeneration

Objectives Positron emission tomography with [18F]fluorodeoxyglucose (FDG-PET) plays a well-established role in assisting early detection of frontotemporal lobar degeneration (FTLD). Here, we examined impact of intensity normalization to different reference areas on accuracy of FDG-PET to discriminate between patients with mild FTLD and healthy elderly subjects.

Methods FDG-PET was conducted at two centers using different acquisition protocols: 41 FTLD patients and 42 controls were studied at center 1, 11 FTLD patients and 11 controls were studied at center 2. All images were normalized to cerebellum, primary sensorimotor cortex (SMC), and cerebral mean (CM). In addition, we applied a recently proposed reference cluster approach to derive a cluster with most preserved FDG uptake in each patients group, in a data-driven manner. To allow for unbiased estimations, a cluster obtained at center 1 was used for normalization of the data from center 2, and the other way round.

Results The centers produced spatially similar clusters with the strongest overlap in anterior parts of cerebellar hemispheres. At center 1, logistic regression analyses based on normalized values from FTLD-typical regions showed that CM, SMC, cerebellar, and cluster normalizations differentiated patients from controls with an accuracy of 76, 78, 85, and 88%, respectively. Corresponding values for center 2 were 85, 79, 79, and 89%. Accuracy provided by cluster normalization was higher (p<0.05) than that of any reference region normalization.

Conclusions Cluster normalization allows an accurate detection of early FTLD-associated metabolic deficits. The established FTLD-specific cluster can be used to improve detection of FTLD on a single case basis at independent centers