Hyperspectral image measurements of skin hemoglobin compared with transcutaneous PO2 measurements

Objective measurements of skin blood flow would have predictive value in assessing the potential for wound healing. In this study, we evaluated the relationship between transcutaneous PO(2) (tcPO(2)) measurements and hyperspectral reflectance spectroscopy measurements of oxygenated hemoglobin (OxyHgb), deoxygenated hemoglobin (DeOxyHgb), total hemoglobin (Sum = OxyHgb + DeOxyHgb), and hemoglobin saturation (Sat = 100 × OxyHgb/Sum). The effect of varying tcPO(2) probe temperatures (37 °C, 41 °C, and 45 °C) was also assessed.

A Hypermed Oxy-Vu system was used for hyperspectral imaging, with measurements performed 2 minutes after removing tcPO(2) probes (Radiometer). Twenty-three sections of foot or wrist skin in four healthy volunteers were measured at 37 °C, 41 °C, and 45 °C using both modalities.

TcPO(2) at 37 °C was 23.1 ± 24.8 mm Hg, increasing to 63.0 ± 27.3 mm Hg at 45 °C. OxyHgb levels increased from 52.4 ± 25.4 at 37 °C to 101.3 ± 23.8 at 45 °C. Linear regression analysis of the HSI data at 37 °C showed a positive correlation between tcPO(2) and OxyHgb (r(2) = 0.35, P = 0.003), tcPO(2) and DeOxyHgb (r(2) = 0.63, P < 0.0001), and tcPO(2) and Sum (r(2) = 0.60, P < 0.0001), but not Sat (r(2) = 0.001, P = 0.92). As the probe temperature increased, the correlations of tcPO(2) with OxyHgb, DeoxyHgb, and Sum became progressively much weaker.

A marked increase in the HSI measurements of OxyHgb in skin exposed to heated tcPO(2) probes was observed, with tcPO(2), Sat, and Sum measurements also observed to increase with temperature. These measurements were influenced by heat inducing vasodilatation in the superficial skin layers. HSI measurements may be clinically useful for measuring wound healing potential, as they correlate with tcPO(2) levels under normal physiological conditions.