474
J. Eur. Opt. Society-Rapid Publ. 22, 48( 2026)
yellow – green – blue – dark blue colormap with logarithmic intensity scaling to accommodate the wide dynamic range of photon density.
Subsequently, source – detector based DS values were computed in % form by summing fluence within a circular region of interest( ROI) of 10 mm radius, centered at the midpoint between the source and detector locations, at successive 1 mm depth increments along the tissue depth axis perpendicular to the surface. A 10 mm radius ROI was selected to balance spatial specificity with adequate sampling volume, consistent with common practice in optical imaging literature. This scale has been shown to capture relevant optical or hemodynamic responses while minimizing contamination from surrounding regions [ 40, 41 ].
P
DSðÞ¼ z ðxyÞ2Rz ðÞ / ðxyzÞ ð Þ 100 %; ð10Þ P x; y; z / xyz
where R( z) denotes the set of voxels within the ROI at depth z. DS versus depth curves were plotted for the mean cadaveric head at each optical source – scalp layer separation. Additionally, detected photon data stored in. jdat format were decoded and used to plot histograms of photon count versus path length for all diffusive photon transport simulations. This post – processing workflow enabled high – resolution spatial and statistical assessment of DS and photon propagation characteristics.
3 Results
Ray – tracing results obtained from the ballistic photon propagation simulations are illustrated in Figure 1 for the mean cadaveric head with a 250, lm diameterskinpore at t = 10 ps. As the vertical position of the skin pore is varied, changes in ray – pore interactions and reflection patterns at the optical source – scalp interface are modeled, with Figures 1b – 1c showing increasingly pronounced reflection features.
The influence of pore geometry on exit trajectories was assessed across 66 combinations of pore diameters and positions. Figures 2a – 2c present the most pronounced photon weight reduction and dispersion behavior within the ballistic transport regime, specifically for the 500 lm diameter pore at vertical positions of( a) 1,( b) 0.5, and( c) 0 mm off-axis relative to the optical source center. As the skin pore approaches the optical source axis, the ejected photon angle shifts and photon weight is reduced. At a 1 mm offaxis position( Fig. 2a), there is no pore – photon interaction; therefore, the photon weights remain at their maximum, clustered, and initial photon angles are preserved. However, as the pore alignment shifts toward 0 mm( Figs. 2b – 2c), both weight reduction and dispersion behavior are observed for the photon groups positioned at specific radial distance intervals.
Spatial fluence color maps for the mean cadaveric head were generated for the same 66 combinations of pore diameters and positions. Additionally, spatial fluence distributions across the eight individual cadaveric heads were simulated for a 500 lm diameter pore at vertical positions
Figure 1. Example COMSOL ray – tracing plots at t = 10 ps for the mean cadaveric head with a 250, lm diameter skin pore, zoomed to the source – scalp interface. Skin pore at vertical( a) �1.5,( b) �0.8, and( c) �0.7 mm positions. The color bar represents normalized photon weights( 0 – 1).
of 1 and 0.5 mm off-axis to provide an anatomical sensitivity analysis. Figure 3 presents isometric views of the spatial fluence for the mean cadaveric head at a 30 mm source – detector separation. The 500 lm pore diameter and vertical positions 1, 0.5, and 0 mm( Figs. 3a – 3c, respectively) were selected to highlight the maximum observable differences in fluence. Initial photon weights of Diffusive Photon Transport Simulations incorporate the final photon weights from the ballistic photon propagation simulations, thereby these spatial fluence color maps represent the localized photon energy distribution. The transition from off-axis to on-axis pore placement( Figs. 3a – 3c) demonstrates a visually observable redistribution of energy density and reduction of photon weights in the vicinity of the source – scalp interface.
DS profiles as a function of depth were generated for all six pore diameters across eleven vertical positions for the mean cadaveric head, alongside comparisons across the eight individual cadaveric models. Figure 4 presents a representative DS profile for the 500 lm diameter pore at a 30 mm source – detector separation. Analysis of the complete dataset( detailed in Supplementary Materials, Section C: Diffusive Photon Transport Simulations, Figures S149 – S156) revealed that the eleven vertical pore positions exhibit virtually identical trends across the entire depth range. This high degree of uniformity indicates that while pore alignment modulates the total energy weight( Fig. 3), the relative spatial distribution of sensitivity is independent of localized surface microtopography. This trend is further corroborated by the comparisons across the eight individual cadaveric heads, which despite measurable variance in decay rates driven by anatomical differences, maintain a consistent baseline sensitivity profile. Together, these observations suggest that the fundamental DS is governed by the