FUNDAMENTAL STUDY OF AUTOMATIC CYTO-SCREENING FOR UTERINE CANCER .1. FEATURE EVALUATION FOR PATTERN-RECOGNITION SYSTEM

Google Scholar
  • 1 January 1977
    • journal article
    • research article
    • Vol. 21  (1) , 72-78
Abstract
A basic study was done to determine the parameters of a pattern recognition system for the automatic assessment of human cytologic cell samples. Various cell features were extracted, whose combinations were evaluated by an ambiguity function. The highest reliability can be obtained with a combination of the features of nuclear staining, nuclear area, area of cytoplasm, nuclear/cytoplasmic ratio, nuclear shape and chromatin pattern. Recognition of the nuclear edge and chromatin patterns is complicated and makes automation difficult. Even if these 2 features are omitted, false positives do not exceed 20%. Screening of abnormal cells can be done by image recognition procedures by the use of a computer.

This publication has 1 reference indexed in Scilit: