目的:研究并建立一种适于生物组织介电特性数学模型快速拟合的新算法。方法:基于单一频散区间内生物组织电阻抗频谱特性在复阻抗平面中为一段圆弧的特性,采用最小二乘算法原理,建立了一种非迭代的快速拟合算法。结果:针对含误差的仿真数据拟合结果表明该方法可将随机误差降低10倍左右,针对生物组织实测数据的拟合结果表明拟合误差小于1.24%,拟合精度高。结论:新方法符合大数据量的生物组织介电特性建模需求。
Objective: To develop a new fast ifting algorithm that is suitable for calculation characteristic parameters of dielectric properties of biological tissues.Methods: Since the electrical impedance spectrum of biological tissues is manifested as an arc in a whole frequency relaxation zone, we developed a new non-iterative fitting algorithm based on least square method to compute the characteristic parameters.Results: By using this new algorithm, the ifting error on simulated data with random noise was reduced about 10 times, and the ifting error on biological tissue measured data was less than 1.24%.Conclusion: The new algorithm has the potential to process massive data of biological tissue dielectric properties.