Improved differentiation between Churg-Strauss syndrome and Wegener's granulomatosis by an artificial neural network

Objective. To examine the operating characteristics of the American College of Rheumatology (ACR) classification criteria for Churg-Strauss syndrome ( CSS) and Wegener's granulomatosis (WG), and to develop and validate improve d criteria for distinguishing CSS from WG. Methods. The ACR classification criteria for WG and CSS were applied to 40 consecutive CSS patients age- and sex-matched with 40 patients with WG. For ty-three clinical, laboratory, and biopsy parameters were assessed. Artific ial neural networks (ANNs) were trained and tested with all 43 parameters ( set A) and with 15 solely clinical parameters documented at the initial man ifestation of the disease (set B). The ANNs were trained with data from the first 27 CSS and 27 WG patients and validated with data from the next 13 c onsecutive CSS and 13 WG patients. To compare the ANNs with established met hods, traditional format and classification tree criteria were generated us ing the same data sets. Results. Fourteen of 40 CSS patients fulfilled the ACR criteria for WG, whi le 4 WG patients met the ACR criteria for CSS. The ANN, in contrast, reliab ly distinguished all CSS cases from WG cases (parameter set A, accuracy 100 %). For parameter set B, the ANN achieved an accuracy of 100% in the traini ng phase and 96% for validation. The newly formulated traditional format an d classification tree criteria reached an accuracy of 81% and 88%, respecti vely. Conclusion. The ACR criteria for WG do not reliably differentiate between C SS and WG (specificity 65%). An ANN, however, could be trained to correctly allocate all but 1 patient on the basis of clinical data. Indeed, the ANN applied in this study proved superior to established methods of classificat ion. We suggest that an ANN may be effectively applied in the classificatio n of systemic vasculitides.