Fig. 1. Two-frequency MTF (DC versus AC) lookup table for rapid calculation of optical parameters (μ_a and μs')

Fig. 2. Portable quantitative burn imager. (a) Device physical map; (b) optical path schematic

Fig. 3. Time distribution of dorsal burns in SD rats

Fig. 4. Data processing flowchart and data visualization interface. (a) Data processing flowchart; (b) data visualization interface

Fig. 5. Pictures of rat burn skin. (a) Before burn (blank group); (b) 4 s group; (c) 12 s group; (d) 24 s group

Fig. 6. Variation of optical parameters of three groups of different burn areas with time under three wavelengths (the parameter value corresponding to each time period is the average value of ROI area, and its standard deviation is also given. Mann-Whitney U-test is used to compare the differences of data, and the level of significant difference is represented by“*”and * represents P<0.05).(a)－(c) Relative change of absorption coefficient of burn area to the area before burn at each time period under R, G, B wavelengths; (d)－(f) relative change of reduced scattering coefficient of burn area to the area before burn at each time period under R, G, B wavelengths

Fig. 7. Systematic clustering. (a) Systematic clustering flowchart; (b) aggregation coefficient line chart

Fig. 8. Systematic clustering results. (a) 10 min after burn; (b) 120 min after burn

Fig. 9. Factor boxplot at different burn stages. (a) 5 min after burn; (b) 10 min after burn; (c) 30 min after burn; (d) 60 min after burn; (e) 90 min after burn; (f) 120 min after burn

Table 1. Systematic clustering classification accuracy

Table 2. Factor loading for optical parameters