Computer-assisted image analysis can aid the prognostication of advanced-stage neuroblastomas.

PURPOSE: The aim of this work was to detect nuclear parameters related to the prognosis of patients with stage III, IV or DS neuroblastomas. METHODS: Histological sections of 25 operation specimens obtained from children with advanced-stage neuroblastomas were subjected to computer-assisted image analysis. Statistical relationships between nuclear descriptors of the tumor cells and patients' clinical outcome were determined. RESULTS: The coefficient of variability of the mean nuclear area the mean nuclear elongation factor, and the mean nuclear averaged Feret diameter of the neuroblastoma cells were ascertained to be discriminators separating high-grade from low-grade tumors. CONCLUSIONS: The histomorphometrically gauged nuclear parameters may help oncologists to assess the prognosis of patients with advanced-stage neuroblastoma.