Fig. 2

Respiration drives CSF flow in the subarachnoid space. Spinal CSF flow was imaged in live rats using a near infrared (NIR) filter on a commercial operating microscope. A Following extensive muscle dissection to expose the bony anatomy from occiput to T2, the CSF tracer indocyanine green (ICG) was infused into the cisterna magna. B The intraoperative view under white light was captured and used to identify the injection site (arrow) and each spinal level (an example of a region of interest is shown in green, delineating C5). C, D The corresponding fluorescent video captured using the NIR filter (Additional file 1: Video S1 and Additional file 2: Video S2) was then used to evaluate fluorescent signal intensity from C2 to T2. Data were collected in real-time over 20 min and measured at 150 s intervals. A representative image from a spontaneous breathing (C) and mechanically ventilated (D) animal are shown. E–G The fluorescence intensity (mean pixel intensity) over the 20 min time-point is shown. E Pooled data compare the effect of positive pressure mechanical ventilation (Control) to spontaneous breathing (SB). Significantly higher fluorescence intensities were measured in SB rats compared with the controls. F Compares hypertensive rats (high MAP) with controls (with low MAP). No statistically significant differences were detected at any time up to 20 min. G Compares tachycardic rats (high HR) with controls (with low HR). Tracer signal intensities between control and tachycardic rats were similar at all time points except at 2.5 min. Two-way analysis of variance (ANOVA) post hoc Bonferroni’s. All error bars are expressed as ± SEM, n = 6 rats. H Heat maps were generated from fluorescence images. In the heat maps, high tracer signal lies towards the yellow, while low signal lies towards the violet end of the spectrum. I The corresponding 3D heat map. The white bar marks the interval between C1 and C2. J A surface plot of tracer signal intensity demonstrates a pulse like wavefront of CSF tracer. K Tracer waves propagated in a caudal direction and their velocities were computed between C1 and C2. In this study, velocity was significantly higher in SB rats compared to controls (*p = 0.04) as well as hypertensive rats (SB vs. BP *p = 0.04). Two-way analysis of variance (ANOVA) post hoc Bonferroni’s. All error bars are expressed as ± SEM, n = 6 rats