TY - JOUR
AU - Ploem, J.
AB - A computer-controlled image-analysis system (LEYTAS) was used to classify 328 cytological urinary specimens automatically into positive and negative. Classification was based on the presence of cells with increased DNA content or high chromatin contrast. These cells are automatically detected by the LEYTAS system and stored in image memories. Fully automated slide classification on the basis of the total number of detected cells resulted in 33% false-positives and 2% false-negatives. Positively classified preparations were then further investigated by rapid visual evaluation of the image memories in order to eliminate any detected artifacts. After this procedure the results were: 12% false-positives and 4% false-negatives. Automated screening of urinary cytology specimens for the detection of bladder cancer, therefore, is feasible. Further analysis of the unbiasedly detected cells for parameters, such as chromatin distribution and nuclear shape, provides a way to gain objective information with respect to grading, therapy, and prognsois.
TI - Results of the automated analysis of 328 bladder specimens using the Leyden television analysis system (LEYTAS)
JF - World Journal of Urology
DO - 10.1007/BF00326747
DA - 2004-08-27
UR - https://www.deepdyve.com/lp/springer-journals/results-of-the-automated-analysis-of-328-bladder-specimens-using-the-xw7madMEy5
SP - 77
EP - 81
VL - 1
IS - 2
DP - DeepDyve
ER -