Ntziachristos et al. respond:

The physics of photon propagation1 in tissues unequivocally supports the notion that planar imaging is surface weighted because signal intensity from deep-seated tumors drops exponentially as a function of depth. Therefore, planar imaging will preferentially detect and localize superficial fluorochrome activity. Indeed, Hoffman and his colleagues2 had to use invasive skin flap windows in previous work to improve on tumor visualization in the lung using planar imaging. Similarly, changes in tumor vascular volume fraction because of angiogenesis alter the internal optical properties of the lesion and nonlinearly modulate the intensity reported, which can obscure quantification. The spectral images presented in our Perspective were provided by Cambridge Research & Instrumentation (CRI), which originally obtained their dual-color mouse tumor model from AntiCancer, the company for which Hoffman is Chairman of the Board. Therefore, our comments regarding the weak fluorescence of this mouse and subsequent spectral analysis reflect back to AntiCancer, which originally provided the mouse images for analysis to CRI.