FUNDAMENTAL STUDY OF AUTOMATIC CYTOSCREENING FOR UTERINE CANCER .4. SAMPLE REQUIREMENTS FOR CYBEST AND SIMULATION TEST OF CELL DISPERSION
- 1 January 1977
- journal article
- research article
- Vol. 21 (4) , 531-535
Abstract
The preparation of samples with good cell dispersion and without cell overlapping is a particularly important problem, whether for image processing systems or flow systems. Simulation tests were done to estimate a cell density with good cell dispersion and a minimum number of overlapping cells. A little less than 50 cells/mm2 was disclosed to show a condition of sufficient cell dispersion. By microscopic observation and counting of the malignant cells using 10 human cases of uterine cervical carcinoma, in most instances over 50 malignant cells were contained in 20,000-40,000 cells on the smears. When the cell density is .apprx. 50/mm2, the scanning area of the CYBEST [Cytobiological Electronic Screening System] of 40 .times. 20 mm may contain up to 40,000 cells. The probability that such a number of cells contains malignant cells is extremely high, in fact, nearly 100%. By field tests, it was experimentally confirmed that these cell densities are adequate for the function of CYBEST.This publication has 3 references indexed in Scilit:
- FUNDAMENTAL STUDY OF AUTOMATIC CYTO-SCREENING FOR UTERINE CANCER .2. SEGMENTATION OF CELLS AND COMPUTER-SIMULATION1977
- FUNDAMENTAL STUDY OF AUTOMATIC CYTO-SCREENING FOR UTERINE CANCER .1. FEATURE EVALUATION FOR PATTERN-RECOGNITION SYSTEM1977
- FUNDAMENTAL STUDY ON AUTOMATIC CYTO-SCREENING FOR UTERINE CANCER .3. NEW SYSTEM OF AUTOMATED APPARATUS (CYBEST) UTILIZING PATTERN-RECOGNITION METHOD1977