Clinicopathological features of fibrosarcomatous dermatofibrosarcoma protuberans and the construction of a back-propagation neural network recognition model

Abstract Background Fibrosarcomatous dermatofibrosarcoma protuberans (FS-DFSP) is a form of tumor progression (DFSP) with an increased risk metastasis and recurrence. Few studies have compared the clinicopathological features FS-DFSP conventional DFSP (C-DFSP). Objectives To better understand epidemiological characteristics FS-DFSP. Methods We conducted cohort study 221 patients diagnosed built recognition model back-propagation (BP) neural network for Results Twenty-six 195 C-DFSP were included. There no differences between regarding age at presentation, diagnosis, sex, size growth. The negative ratio CD34 in (11.5%) was significantly lower than that (5.1%) ( P = 0.005). average Ki-67 index (18.1%) cases higher (8.1%) < 0.001). classification accuracy BP training samples 100%. correct rates misdiagnosis 84.1% 15.9%. Conclusions clinical manifestations are similar but large immunohistochemistry. feasibility high