Abstract:Background and Objectives: The possibility of restoring microhemodynamics on the human body surface was investigated through the analysis of skin temperature oscillation signals using integer and fractional order derivatives. Materials and Methods: Microhemodynamic data were simultaneously recorded via photoplethysmographic imaging and infrared thermography in four regions of the hand. To reconstruct microhemodynamics from temperature data, a previously described thermal wave model, along with integer and fractional order derivatives, were applied. A comparative analysis of the amplitude-frequency and phase-frequency characteristics of these transformations was conducted. The fractional-order derivative of temperature oscillations was calculated as the Riemann – Liouville differintegral. For a group of subjects, correlations were computed between the reconstructed microhemodynamic results during a 15-minute resting state using the thermal wave model and the integer/fractional-order derivatives of temperature. Results: It has been established that employing a fractional-order derivative of order 0.4 has yielded the best correlation between the frequency characteristics and those of the thermal wave model. The enhanced temporal-domain signal correlation achieved with the fractional-order derivative, compared to the integer-order derivative, is attributed to more accurate amplitude-frequency and phase-frequency transformations of temperature oscillations. These transformations align with the attenuation and dispersion processes of thermal waves in the skin. Conclusions: For precise restoration of microhemodynamics using skin temperature time derivatives, a the fractional-order derivative of 0.4 is preferable over integer-order derivatives. The described method can serve as a thermalbased technique for investigating blood flow oscillations in microvessels across multiple anatomical regions simultaneously.
Fulltext:
PDF file (1655 kB)