Characterisation of the BOLD response time course at different levels of the auditory pathway in non-human primates

Abstract

Non-human-primate fMRI is becoming increasingly recognised as the missing link between the widely applied methods of human imaging and intracortical animal electrophysiology. A crucial requirement for the optimal application of this method is the precise knowledge of the time course of the Blood Oxygenation Level Dependent (BOLD) signal. We mapped the BOLD signal time course in the inferior colliculus (IC), medial geniculate body (MGB) and in tonotopically defined fields in the auditory cortex of two macaques. The results show little differences in the BOLD-signal time courses within the auditory pathway. However, we observed systematic differences in the magnitude of the change in the BOLD signal with significantly stronger signal changes in field A1 of the auditory cortex compared to field R. The measured time course of the signal was in good agreement with similar studies in human auditory cortex but showed considerable differences to data reported from macaque visual cortex. Consistent with the studies in humans we measured a peak in the BOLD response around 4 s after the onset of 2-s broadband noise stimuli while previous studies recorded from the primary visual cortex of the same species reported the earliest peaks to short visual stimuli several seconds later. The comparison of our results with previous studies does not support differences in haemodynamic responses within the auditory system between human and non-human primates. Furthermore, the data will aid optimal design of future auditory fMRI studies in non-human primates.

Fig. 1

Schema of the ‘sparsed’-fMRI design with stimulus jitter. Two sound stimuli of different durations were presented in separate sessions. The onsets of the stimuli preceded the data acquisition by different time periods from 0 to 11 s to sample the time course of the BOLD response. Data acquisitions with stimulus presentation were interleaved with acquisitions without stimulus to obtain baseline trials.

Fig. 2

T-maps of the broadband noise vs. silence contrast for the two monkeys. T-maps are displayed in panel A for monkey C and in panel C for monkey W. The approximate location of the displayed slices I–VII is shown in panel B. Borders of the individually identified auditory fields are outlined in blue. For a detailed organisation of the fields see Fig. 3. A1: primary auditory cortex, R: rostral auditory field, MGB: medial geniculate body, IC: inferior colliculus, A: anterior, P: posterior, L: lateral.

Fig. 3

Tonotopic organisation of the auditory cortex. Panels A and D display the t-value difference map for high frequency bandpass noise (Hf) compared to low frequency bandpass noise (Lf) in the two monkeys C and W. A map of the voxels that give the strongest response to one of three frequency bands Hf, Mf and Lf is displayed in panels C and F. Panels B and E show the location of mirror reversed tonotopic gradients (positive: low to high from posterior to anterior; negative: high to low from posterior to anterior. The borders of the identified fields are outlined in black. G shows the organisation of auditory fields according to Hackett et al., 2001 and Petkov et al., 2006. Hf: high frequency band (8–16 kHz), Mf: middle frequency band (2-4 kHz), Lf: low frequency band (0.5–1 kHz), A1: primary auditory cortex, R: rostral field, RT: rostrotemporal field, CM: caudomedial field, CL: caudolateral field, MM: middle medial field, ML: middle lateral field, RM: rostromedial field, AL: anterolateral field, RTM: rostrotemporomedial field, RTL: rostrotemporolateral field. Tpt: temporoparietal junction.

Fig. 4

BOLD response time courses for short and sustained stimuli from four different areas in the auditory pathway. A: top row shows average responses to 2-s stimuli; bottom row shows responses to 8-s stimuli. First column: monkey C, second column: monkey W. Bars represent standard error. B: time courses of the individual sessions. Same arrangement as in A. C: measured BOLD signal time-course from A1 after 2-s stimulus (red, solid line), 8-s stimulus (blue, solid line), the prediction of the 8-s stimulus response based on the 2 s data (blue dotted line) and difference between measured and predicted time course (black line).