Title :
Segmentation of cell images using an expert system
Author :
Song, Loling ; Donovan, Richard M.
Author_Institution :
Sch. of Med., California Univ., Davis, CA, USA
Abstract :
The objective of this work was to develop a rule-based expert system for recognition of single or overlapping cells. A local threshold was found automatically by gradient operations and then used for chain-code formation. The shape analysis was done on the chain code of cell edges. The curvature function was derived from the chain code. The rule-based expert system assisted in analyzing the curvature function and making decisions on object recognition. The preliminary results in this study demonstrate the potential of expert systems in cell image segmentation.<>
Keywords :
expert systems; medical diagnostic computing; AIDS patients´ lymphocytes; cell images segmentation; chain-code formation; curvature function; medical expert system; overlapping cells; rule-based expert system; single cells;
Conference_Titel :
Engineering in Medicine and Biology Society, 1988. Proceedings of the Annual International Conference of the IEEE
Conference_Location :
New Orleans, LA, USA
Print_ISBN :
0-7803-0785-2
DOI :
10.1109/IEMBS.1988.95182
