Many challenges in plastic and reconstructive surgery are multi-tissue， multi-scale， and strongly timedependent. Traditional cell and animal models are hampered by limited human relevance， insufficient control of mechanical loading， limited fidelity in physicochemical emulation， and the lack of noninvasive functional assessments. Microphysiological systems integrate controlled microfluidic perfusion， programmable stretch， chemical gradients， and human-derived multicellular co-culture to reconstruct key processes on a single platform： In wound-healing studies， transepithelial electrical resistance and permeability assessments quantitatively relate wound-closure rate to barrier restoration； In keloid disease modeling， programmable tensile loading and stable TGF- β gradients enable dissection of extracellular-matrix remodeling and macrophage polarization during fibrosis； For fat grafting， an “endothelium high-shear—membrane-separated exchange—adipose low-shear” architecture quantifies oxygenation， nutrient delivery， and the immunometabolic state； For neural and musculoskeletal regeneration， topographic cues， chemical concentration gradient， and electrical/mechanical stimulation drive directed maturation and support functional evaluation. This review synthesizes these model elements and representative applications， outlines engineering implementations （air-liquid interface， cyclic stretch， membrane-separated dual-chamber designs， and multimodal stimulation）， and discusses key challenges （crossplatform comparability， donor-to-donor variability， and long-term operational stability）—to provide a practical roadmap for deploying advanced in vitro models in preclinical validation and individualized preoperative assessment.

Microphysiological systems, Wound healing, Keloids, Fat grafting, Regenerative medicine, In vitro