作者: C. J. Cornelisse , A. M. J. Driel-Kulker , F. Meyer , J. S. Ploem
DOI: 10.1111/J.1365-2818.1985.TB02566.X
关键词: Nuclear atypia 、 Image transformation 、 Breast cancer 、 Cytology 、 Mathematical morphology 、 Segmentation 、 Radiology 、 Pathology 、 Nuclear area 、 Biology 、 Cytophotometry
摘要: In order to develop an objective grading system for nuclear atypia in breast cancer, image analysis technique has been applied the automated recognition of enlarged and hyperchromatic nuclei cytology specimens. The segmentation algorithm, based on 'top hat' transformation developed mathematical morphology, is implemented LEYTAS microscope system. performance algorithm evaluated fifty malignant eighty-five benign lesions by visual inspection displayed 'flagged' objects. mean number flagged objects per 1600 fields cancers was 887 (range 0-7920) which 87% consisted single, atypical nuclei. For 30 0-307) 20% were single By adaptation parameter values, a more extreme subpopulation could be discriminated. large interspecimen variation cancer results related differences DNA content distribution area, determined independently with scanning cytophotometry, some extent histological type.