Automated nuclear segmentation in the determination of the Ki-67 labeling index in meningiomas

Clin Neuropathol. Mar-Apr 2006;25(2):67-73.


Objective: Assessing the Ki-67 labeling index (LI) is laborious and time consuming. Therefore, an automated computer-based method was developed, which is able to identify and analyze immunolabeled and hematoxylin-stained nuclei in digital images of routine immunohistochemical slides.

Material and methods: The method is based on a plugin for the public domain image analysis software ImageJ, which runs on every operating system (free download at Percentage of Ki-67 immunostained nuclei were determined in 5 high power fields (x40) of immunostained slides (DAB detection technique, hematoxylin counterstain) of 20 Grade I, 20 Grade II, and 10 Grade III meningiomas conventionally by two independent investigators and automatically, respectively. The time effort was measured for each counting procedure.

Results: Enumerating conventionally or automatically did not reveal any significant differences in the mean labeling indices. Ki-67 LIs discriminated sufficiently between meningiomas of Grade I (median 1.7% Investigator 1 and 1.5% Investigator 2 vs. 1.5% automatically), Grade II (7.6%, 8% vs. 7.3%), and Grade III meningiomas (22%, 21% vs. 22%). The computer-based results correlated very closely with those obtained by manual counting (correlation coefficient = 0.98). The mean time effort for counting procedure per image was 374 s (130 s-435 s) for the conventional and 11 s (7 s-12 s) for the automated method.

Conclusions: The described method can reliably assess the Ki-67 LI much faster than conventional enumerating. The computerized method has the advantages of objectivity, accuracy, repeatability, and ease of use. There is no request for special stains nor special image acquiring systems. The plugin can be downloaded at the "Morphometrie" section of

Publication types

  • Comparative Study
  • Validation Study

MeSH terms

  • Biomarkers, Tumor / analysis*
  • Cell Nucleus / pathology*
  • Humans
  • Image Processing, Computer-Assisted
  • Immunohistochemistry / methods*
  • Ki-67 Antigen / metabolism*
  • Meningeal Neoplasms / metabolism*
  • Meningeal Neoplasms / pathology
  • Meningioma / metabolism*
  • Meningioma / pathology
  • Reproducibility of Results
  • Sensitivity and Specificity


  • Biomarkers, Tumor
  • Ki-67 Antigen