Grading of urinary bladder tumors by automated microscopic image analysis

Studies were undertaken on biopsies from urinary papillomas and urothelial carcinomas of different grades (G1 to G3) along with normal urinary bladder mucosa from a total of 220 patients. They were performed by use of automated microscopic image analysis (system Robotron A 6471, software AMBA/R). As a result, the tumor grading was improved and lead to the separation of both papillomas and G1-carcinomas with high proliferative activity. In addition, it was possible to divide the G2-carcinomas into 2 groups which probably show a different biological behaviour.