Determination of Volume-weighted Mean Nuclear Volume by Automatic Stereology

Malignant cells that spend the majority of their time in synthesis (S) phase have mean nuclear volumes (MNV) that are significantly larger than non-malignant (normal) cells.

Furthermore, cancers with higher MNV tend to show greater malignancy and earlier metastases. Here we apply this approach to automatic quantification of MNV of cervical biopsies for the purpose of disease staging.

For these types of cancers, MNV provides an important stereology-based method for separating malignant cells from normal cells. Here we show a fully automatic three-step procedure procedure for quantifying MNV and assessing the malignancy potential of cervical cancer cells (Chaudhury et al. 2013).

In the first step, images of H&E-stained cervical tissue sections are segmented to isolate cervical cell nuclei (not shown). Second, point-sampled intercepts (PSI) are used to sample segmented nuclei in a volume-weighted manner (Figure 1 below). Because PSI preferentially samples larger nuclei for the MNV estimate, Gundersen and Jensen (1985) selected the term “volume-weighted mean nuclear volume (PSI-Vv)” for this estimator. The final step is collection of line lengths (l) across sampled nuclei and computation of PSI-Vv.

Figure 1. Point sampled intercepts (PSI) for selecting nuclei in a volume-weighted manner.

volume-weighted mean nuclear volume

Figure 2. The nucleator method for quantifying mean nuclear volume of PSI-sampled nuclei.

12- 4 Nucleator.Prin2 copy

Gundersen, H.J. and Jensen, E.B. Stereological estimation of the volume-weighted mean volume of arbitrary particles observed on random sections. Journal of Microscopy, 138: 127-142, 1985.

Chaudhury, B., Phoulady, H.A., Goldgof, D., Hall, L.O., Mouton, P.R, Hakim, A., Siegel, E.M. An Ensemble Algorithm Framework for Automated Stereology of Cervical Cancer, Proceedings of the International Conference on Image Analysis and Processing (ICIAP), Springer (Berlin) Vol I, pp. 823-32, 2013.