Depth-four threshold circuits for computer-assisted X-ray diagnosis?

Abstract

The paper continues our research from [1]. We present a stochastic algorithm that computes threshold circuits designed to discriminate between two classes of CT images. The algorithm is evaluated for the case of liver tissue classification. A depth-four thresh- old circuit is calculated from 400 positive (abnormal findings) and 400 negative (normal liver tissue) examples. The examples are of size n = 14161 = 119×119 with an 8 bit grey scale. On test sets of 100+100 examples (all different from the learning set) we obtain a correct classification of about 96%. The classification of a single image is performed within a few seconds.