Abstract
Perioperative identification of ischemic-hypoxic tissue is essential for reducing the risk of mastectomy skin flap necrosis (MSFN), a major complication following mastectomy. Implant-based breast reconstruction is jeopardized in the presence of MSFN. However, existing technologies are limited by depth insensitivity, dye-related risks, the need for contact sensors, and an inability to simultaneously assess both blood flow and oxygenation. A multi-wavelength speckle contrast diffuse correlation tomography (MW-scDCT) system was developed for noncontact, dye-free, and depth-sensitive imaging of tissue hemodynamics in a rat model incorporating four distinct flap scenarios: sham (SH), implant (IM), half necrosis (HN), and full necrosis (FN), representing varying degrees of tissue viability. MW-scDCT measured longitudinal changes in relative blood flow index ( rBFI ) and oxy- and deoxy-hemoglobin concentrations ( Δ[HbO2] and Δ[Hb] ) over seven days. Repeated measures analysis of variance and quadratic discriminant analysis were employed for statistical analysis. MW-scDCT enabled longitudinal imaging of tissue blood flow and oxygenation distributions, revealing significant differences across flap types and over time. In terms of discriminative power, integrating rBFI , Δ[HbO2] , and Δ[Hb] into a multivariable classification model substantially improved accuracy, achieving 80-95% compared to 40-70% using individual parameters. Binary classification of necrotic (FN + HN) versus non-necrotic (SH + IM) flaps showed up to 50% improvement in accuracy gains. MW-scDCT effectively distinguished necrotic from viable flaps and provided critical early postoperative insights into MSFN. These findings support its potential clinical utility as a perioperative monitoring tool to guide decision-making, reduce flap failure risk, and improve reconstructive outcomes.