Urinary Bladder Tumor Grade Diagnosis Using On-line Trained Neural Networks

This paper extends the line of research that considers the application of Artificial Neural Networks (ANNs) as an automated system, f or the assi nment of tumors rade. One hundred twenty nine cases were classified accordin to the WHO radin system by experienced patholoists in three classes: Grade I, Grade II and Grade III. 36 morpholo ical and textural, cell nuclei feig res represented each case. Thesef eatures were used as an input to the ANN classifier, which was trained usin a novel stochastic trainin al orithm, namely, the Adaptive Stochastic On-Line method. The resultin automated classification system achieved classification accuracyof 90%, 94.9% and 97.3%f. tumors of Grade I, II and III respectively.