MEDIAL AXIS TRANSFORM-BASED FEATURES AND A NEURAL-NETWORK FOR HUMAN-CHROMOSOME CLASSIFICATION

Medial axis transform (MAT) based features and a multilayer perceptron (MLP) neural network (NN) were used for human chromosome classificati on. Two approaches to the MAT, one based on skeletonization and the ot her based on a piecewise linear (PWL) approximation, were examined. Th e former yielded a finer medial axis, as well as better chromosome cla ssification performances. Geometrical along with intensity-based featu res were extracted and tested. The probability of correct training set classification of five chromosome types was 99.3-99.6%. The probabili ty of correct test set classification was greater than 98% and greater than 97% using features extracted by the first and second approaches, respectively. It was found that only 5-10, out of all the considered features, were required to correctly classify the chromosomes with alm ost no performance degradation.