Figure 8.

Schematic representation of the image processing workflow of predicting PDT treatment response from 3D ultrasound (US) and PA images. (A) Example slice of combined PA and US image of tumor anatomy and StO₂ content, respectively. (B) 3D StO₂ overlaid on tumor anatomy map, and (C) US‐imaged tumor anatomy alone. Each StO₂ value of a given region in a slice (A) is then fed into D: the computational pipeline for determining StO₂ variability and thus treatment responsiveness. Should the StO₂ values satisfy the requirements of both the 6‐h and 24‐h post‐PDT cutoffs, a Boolean value corresponding to red (not responsive to treatment) or green (responsive to treatment) is then used to reconstruct a 3D map of regional tumor responsiveness (E). Adapted with permission from 73.