Table 1. Microvascular network architecture quantitative metrics.
Metric | Unita | Description | Potential biological relevance |
---|---|---|---|
Vessel area density (VAD) | % | Perfused blood vessel area in binarized OCTA MIP image divided by total image area | Indicates utilization of microvessels. Increased VAD is associated with angiogenesis |
Vessel length density (VLD) | % | Total length of perfused vessels measured along the vessel centerline divided by the total image area. Length measurements are calculated from the skeletonized OCTA MIP image | Indicates oxygenation/nutrient delivery dysfunction. Increased VLD is associated with angiogenesis |
Average and distribution of vessel diameter | μm | Diameter measurements acquired using a local thickness algorithm in ImageJ applied to the binarized OCTA MIP | Reveals information about dilation and regression; the distribution allows further differentiation between changes in diameter and number of perfused vessels |
Average and distribution of vessel length | mm | Length of each identified vessel segment along the centerline of the vessel calculated from the skeletonized OCTA MIP image | Network interconnectivity and branching patterns indicates oxygenation/nutrient delivery dysfunction |
Average and distribution of tortuosity | 1 | Segment length along the centerline divided by chord length for each identified vessel calculated from the skeletonized OCTA MIP image | Increased blood vessel tortuosity can indicate pathological microvascular remodeling and/or ischemia |
Branchpoint density | Nodes/mm | Number of identified nodes divided by total vessel length | Network interconnectivity-based indicator of resistance to blockage or occlusion and flow dynamics |
Fractal dimension | 1 | Indication of how the network fills space on variable length scales, calculated using the box counting method | Indicator of branching architecture within the network; indicator of tortuosity and microvascular remodeling |
aA unit of 1 indicates a dimensionless metric.