Figure 3

Co-registered macro DCE-MRI to micro svOCT vascular correlations. (A) \({T}_{2}\)-weighted structural MRI scan of the window chamber with tumor delineated by the red contour. (B) \({k}_{trans}\) parameter map of the tumor, averaged over two depth slices (total depth of 1 mm to correspond with svOCT’s imaging penetration), in units of min⁻¹ indicated by the colour bar. (C) svOCT segmented depth-encoded vasculature coregistered to (B). The grey dotted line in (B) and (C) shows one position of the \(1 \space{\mathrm{mm}}^{3}\) sliding window VOI with numbered edges that correspond to the number locations in (D) and (E). The various semi-quantitative and quantitative MR vascular metrics in the resulting DCE-MRI voxels (8 \({k}_{trans}\) voxels in this example) (D) are directly compared to microvascular biomarkers derived from the corresponding svOCT 3D microvascular map (E). The VOI then slides throughout the delineated tumor contour, with such analysis repeated at all positions. (B)–(E) were generated using MATLAB R2020A software (MathWorks, Inc., Natick, MA, USA).