Flow, Resistance, and Turbulence: Elucidating the Hemodynamics of Free Flaps

BACKGROUND: The hemodynamics of free flaps remain incompletely characterized. Free flaps introduce nonphysiologic inflow conditions, where abrupt geometric changes can generate reflected waves that distort local flow. Fourier-domain analyses can isolate these reflections to quantify alterations in flow conditions. This study characterizes the hemodynamics of free flaps by linking flow, resistance, turbulent waveform changes, and intrinsic flap characteristics. METHODS: A retrospective review was conducted for patients who underwent free tissue transfer. For each flap, transit-time flow data, including flow and pulsatility index (PI), as well as arterial and venous pedicle calibers, were recorded intraoperatively. A short-time Fourier transform was applied to sequential segments of each flow signal. Harmonic distortion (HD), defined as the ratio of the cumulative power of the first n harmonics to the power at the fundamental, was calculated for the first harmonic (HD1), first five harmonics (HD5), and first ten harmonics (HD10). RESULTS: Seventy free flaps in 51 patients were analyzed. Across all flaps, mean arterial inflow was 9.0 mL/min (SD=5.9) and mean PI was 3.8 (SD=3.8). Flow was lowest in ALT flaps, intermediate in DIEP flaps, and highest in LD flaps. PI showed the inverse pattern. ALT flaps exhibited the greatest HD; DIEP flaps had the lowest HD at low-order harmonics but converged with LD flaps at high orders. Arterial pedicle caliber did not correlate significantly with hemodynamic metrics (P>0.05), whereas larger venous diameter correlated with higher flow, lower PI, and lower HD (P<0.05). Flow and PI showed a strong inverse relationship (P<0.001). CONCLUSION: Free flap perfusion reflects a dynamic interplay among flow, resistance, and turbulence. PI and harmonic distortion provide markers of resistance and waveform complexity. Venous caliber, not arterial, correlates with favorable hemodynamic profiles. Flap spectral signatures may enable functional characterization beyond anatomic descriptors.