Fig. 3. Schematic outlining the strategy to improve the accuracy and precision of the D map, and its further use to correct the SRRF image.

The untreated D_L(2) map, TIRF image and SRRF image of a CHO-K1 cell expressing Lifeact-EGFP serve as the starting files. First, we create an intensity filtered TIRF mask which is then applied to the D_L(2) map and SRRF image to remove the contributions from off-fibre artefacts generating the $D_{\mathrm{L},\mathrm{on}\left(2\right)}^{\mathrm{TIRF}}$ and ${\mathrm{SRRF}}_{\mathrm{L}}^{\mathrm{TIRF}}$ images respectively. The $D_{\mathrm{L},\mathrm{on}\left(2\right)}^{\mathrm{TIRF}}$ map is then filtered by thresholding D values to remove the slow clusters. The resulting $D_{\mathrm{L},\mathrm{on}\left(2\right)}^{\mathrm{TIRF}}$ map is used as a mask to remove artefacts from the ${\mathrm{SRRF}}_{\mathrm{L}}^{\mathrm{TIRF}}$ image to generate the ${\mathrm{SRRF}}_{\mathrm{L}}^{\mathrm{TIRF},D}$ image. For the dual-colour measurements, the Lifeact SRRF image is filtered only by using the TIRF image as a mask. The orange box indicates an area that contained artefacts in the original SRRF_L image and are now removed in the ${\mathrm{SRRF}}_{\mathrm{L}}^{\mathrm{TIRF},D}$ image.