Physiological noise characteristics in fMRI of the rodent at 11.7T

Introduction Temporal signal-to-noise ratio (tSNR) is the essential parameter for any BOLD fMRI protocol, as it limits the sensitivity to BOLD signal changes, whether they stem from stimulus-induced neuronal activity or from resting-state fluctuations. Thermal noise affects the tSNR as well as the single image SNR (SNR0), but the limiting factor of tSNR may be of different origin: Physiological noise stemming from local T2* fluctuations may severely compromise the tSNR and linearly scales with the signal amplitude. According to the physiological noise model introduced by Krüger and Glover [1], the total noise consists of thermal noise σ0 and the signal dependent physiological noise σp as descri bed by Eq. (1). Thus, gains in SNR0 above a certain level will no longer translate into an improved temporal SNR. This saturation effect between tSNR and SNR0 is expressed in Eq. (2). Animal imaging profitted immensely from higher field strengths and improved hardware, especially in the field of fMRI. The goal of this study is therefore to investigate the physiological noise characteristics in BOLD fMRI protocols of the rodent at a very high field strength of 11.7 Tesla. In particular, the individual contributions of thermal and physiological noise shall be determined using the physiological-to-thermal-noise ratio (PTNR) as defined in Eq. (3).