METHOD FOR PREDICTION OF BREAST CANCER USING ATTENTIVE HIERARCHICAL ADAPTIVE NEURO-FUZZY INFERENCE SYSTEM, RECORDING MEDIUM AND DEVICE FOR PERFORMING THE METHOD

摘要

A method for predicting breast cancer using a hierarchical adaptive neural fuzzy inference system includes: inputting a variable set calculated from a digitized image of a microneedle aspirate (FNA) of breast mass data; Extracting a function using different A-ANFIS parameters from each A-ANFIS through a first layer of AH-ANFIS (Attentive Hierarchical ANFIS) consisting of A-ANFIS (Attentive ANFIS) regression units; Applying the proliferation concentration mechanism of the second layer of AH-ANFIS to a plurality of functions extracted from the first layer of the AH-ANFIS, and selecting activated rules and critical variables for predictable interpretation; Determining a recurrence of the patient by selecting a maximum reliability from the measured output probability based on the selected activated rule and the threshold variable; And displaying and outputting the patient's breast cancer classification result as malignant or benign. Accordingly, the interpretability of breast cancer prediction can be improved.