THG microscopy in the mouse ear.

<p>(a,b) Intravital imaging in different animals with arteriole (A), venule (V) and nerve fiber (N) running in parallel. In addition to the THG signal (cyan) SHG (red) is displayed. The arrow points to a typical linear vessel wall signal found only in arterioles. Scale bars 20 µm. The declining THG signal strength in more axial parts of a vessel are easiest to recognize in the arteriole in (a) and the venule in (b). (c–e) Flow velocity profile measurements by THG shifted line scan in the venule shown in b. (c) Clippings from x-t-representations of 12 parallel scan lines spaced 4.4 µm from each other. Lines 1, 11 and 12 were located in the vessel wall and thus show no movement. Note the shallower angles (faster velocity) and decreased intensity in the central scans. (d) Measurements derived from c by manual evaluation of individual streaks (circles) with mean values (horizontal lines) and mean values from LS-PIV calculations (diamonds). Scan lines 2 (0 µm) to 10 (35.2 µm) from (c) are included. Manually measured velocities revealed significant differences between the scan lines (p<0.0001, ANOVA) with significant differences between all direct neighbors (p<0.05 or smaller, post hoc Newman-Keuls test). For scan line 2 (0 µm), only a part of the x-t-representation showed blood flow. Apparently this line was close to the vessel wall so that slight movement could shift it outside. No reasonable LS-PIV average could thus be obtained. LS-PIV evaluations of the other scan lines confirmed significantly different blood flow velocities in general (p<0.0001; ANOVA, n = 88 for each line) and for all direct neighbors (****), except for the two central lines. (e) Measurements by manual evaluation for the arteriole shown in (b) Velocity measurements were significantly different between 0 and 4.4 µm lines (p = 0.0075) and between 8.8 and 13.2 µm lines (p = 0.0129). (f) LS-PIV results for the same arteriole. Only systolic maxima and diastolic minima (circles) and the mean values over all 3.2-millisecond-spaced 179 measurements (line) are shown. In the first scan line systole and diastole could not be identified reliably, therefore only the mean is given. Velocity differences between neighboring scan lines were all significant (p<0.0001) except between 13.2 and 17.6 µm (p = 0.87).