CONTINUOUS CLASS PATTERN-RECOGNITION FOR PATHOLOGY, WITH APPLICATIONSTO NON-HODGKINS FOLLICULAR LYMPHOMAS

Continuous class pattern recognition is a new analytic method that dif fers from existing techniques by recognizing and exploiting the contin uous relationship among classes along diagnostic scales. Continuous cl ass pattern recognition was applied to a wide variety of image process ing features extracted from lymph node biopsy images that were digitiz ed using a Coulter diff3/50 automated research microscope. The resulta nt classifiers correctly subtyped 89% of a set of 37 follicular lympho mas, compared to individual pathologist rates that ranged from 57% to 81%. This study demonstrates that continuous class pattern recognition can significantly reduce this diagnostic error rate. Copyright (C) 19 96 Pattern Recognition Society.