The application discloses a
breast reconstruction type recommendation
system and method based on multi-
modal data, relates to the technical field of
breast reconstruction type recommendation, and aims to solve the technical problems of traditional operation type recommendation depending on linear superposition evaluation of a single index, being unable to quantize the nonlinear complex mechanism of metabolic and vascular injury, and leading to inaccurate
risk evaluation and lack of targeted recommendation, the
system comprising a feature calculation and analysis module, wherein the feature calculation and analysis module comprises a
clinical information structured unit, a metabolic and vascular injury index calculation module, an image feature unit and a test feature unit. Through the metabolic and vascular injury index calculation module, the
nonlinear normalization function, the
standardization function, the interactive
effect analysis algorithm and the
Sigmoid function are integrated, the true risk difference between patients with short
disease duration but poor blood
glucose control and patients with long
disease duration but good blood
glucose control is effectively distinguished, the
risk evaluation is more in line with the pathophysiological law, and a scientific quantitative basis is provided for operation
type selection.