Computer-Based Diagnostic Analysis of Cells in the Urinary Sediment

Evaluation of images of cells from the urinary sediment by a computer-based microscope has yielded important preliminary results. Computer identification of benign and malignant urothelial cells can now be accomplished with a small margin of error. Furthermore, 2 categories of atypical urothelial cells can be identified. This information has been applied successfully to the study of 12 patients with bladder cancer, including 3 with non-papillary carcinoma in situ. Computer-generated diagnostic profiles based on a relatively modest number of urothelial cells (60 to 150 per patient) proved to be of diagnostic value. It is anticipated that within a few years a practical computerized system of identification of cells in the urinary sediment can be made available for general laboratory use.