Identification of cutaneous perforators for microvascular surgery using hyperspectral technique – A feasibility study on the antero-lateral thigh

Abstract

Aim of the study was to compare perforator vessel location using color-coded Doppler ultrasound and hyperspectral imaging in the area of the antero-lateral thigh. In a cross-sectional case-control study, the bilateral antero-lateral thigh region was examined for perforator vessel location via color-coded Doppler ultrasound (control) and hyperspectral imaging (test). For hyperspectral imaging, all measurements were conducted without cooling (T0) and after 1 (T1), 2 (T2) and 3 min (T3) of cooling. Additionally, in the reperfusion period after cooling, hyperspectral imaging was conducted at 1, 2 and 3 min (T4/T5/T6). Results from color-coded Doppler ultrasound and hyperspectral imaging were matched at all time points (T0–T6). In total, 71/73 perforator vessel locations could be matched (sensitivity: 97%). Matching of color-coded Doppler ultrasound and hyperspectral imaging was significantly influenced by the cooling protocol and the highest matching values were seen at T3 (3 min cooling; 60 perforator vessels) and T4 (3 min cooling & 1 min reperfusion; 62 perforator vessels) without significant differences (sensitivity 98%; p = 0.9). There were significant differences between T4 and T0, T1 (both p < 0.001), T5 (p = 0.045) and T6 (p = 0.012). For clinical proof of concept, a patient case using a free antero-lateral thigh flap for reconstruction of a facial defect after perforator vessel identification via color-coded Doppler ultrasound and hyperspectral imaging (3 min cooling & 1 min reperfusion) was carried out successfully. In conclusion, hyperspectral imaging potentially offers an additional opportunity for non-invasive, user-independent perforator-site assessment if prior cooling of the site is conducted.