A measuring method of effects of herbs based on classification used for intelligent calculation of efficacies of prescriptions

It is important to translate dosage of each medical herb in a prescription into intensity of its effects measured by mathematical models. In this paper, we present intelligent calculational algorithms on efficacies of prescriptions of Chinese Medicine (CM), study “how curative effects of any herb depend on dosage” when constructing CM intelligent systems, and discuss relevant issues and definitions. In order to measure intensities of effects of all herbs reasonably, the natural classification method is proposed to classify herbs into 38 natural classes according to their dosage ranges using the knowledge of CM dosage and distribution rules of herbs. Then the 38 natural classes are clustered to 6 basic classes reflecting characteristic attributes of various herbs. Based on the results of classified herbs, a set of nonlinear measuring functions of intensities of effects is designed for various basic classes. These functions have been embedded in our intelligent calculation system of efficacies of prescriptions. The effects of 560 herbs and curative efficacies of more than 1000 prescriptions derived from “ShangHan Lun”, “JinGui YaoLue”, “Ming Jia Yi An”, etc., have been calculated and discovered successfully. It is proved that the accuracy of theses results by these models has been obviously enhanced.