On semiautomatic estimation of surface area.

In this paper, we propose a semiautomatic procedure for estimation of particle surface area. It uses automatic segmentation of the boundaries of the particle sections and applies different estimators depending on whether the segmentation was judged by a supervising expert to be satisfactory. If the segmentation is correct the estimate is computed automatically, otherwise the expert performs the necessary measurements manually. In case of convex particles we suggest to base the semiautomatic estimation on the so-called flower estimator, a new local stereological estimator of particle surface area. For convex particles, the estimator is equal to four times the area of the support set (flower set) of the particle transect. We study the statistical properties of the flower estimator and compare its performance to that of two discretizations of the flower estimator, namely the pivotal estimator and the surfactor. For ellipsoidal particles, it is shown that the flower estimator is equal to the pivotal estimator based on support function measurements along four perpendicular rays. This result makes the pivotal estimator a powerful approximation to the flower estimator. In a simulation study of prolate and oblate ellipsoidal particles, the surfactor also performs well for particles which are not extremely elongated. In particular, the surfactor is not very much affected by the singularity in the surfactor formula or by possible inaccuracies in the necessary angle measurements. We also assess the performance of the semiautomatic procedure in a study of somatostatin positive inhibitory interneurons from mice hippocampi. Only 35% of the cells needed to be analysed manually and an important decrease in workload was obtained by using the semiautomatic approach.

The semi-automatic nucleator.

The nucleator is a well-established manual stereological method of estimating mean cell volume from observations on random cell transects through reference points of the cells. In this paper, we present an automated version of the nucleator that uses automatic segmentation of the boundaries of the cell transects. An expert supervises the process. If the segmentation is judged to be satisfactory, an estimate of the cell volume is calculated automatically on the basis of the whole cell transect. In the remaining cases, the expert intervenes and uses the classical nucleator. The resulting estimator is called the semi-automatic nucleator. In this paper, we study the statistical properties of the semi-automatic nucleator. Formulae for the bias and mean square error are derived. The semi-automatic nucleator may have a small bias but will still in most cases be more efficient than the classical nucleator. Procedures for estimating bias and mean square error from a pilot study are provided. The application of the semi-automatic nucleator is illustrated in a study of somatostatin positive inhibitory interneurons which were genetically labelled with green fluorescent protein (GFP). The cells were sampled with an optical disector. The centre of mass in a central cell transect was used as reference point. It is found in this study that the number of cells needed for obtaining, for instance, a 5% precision of the estimate of mean cell volume is 150 and 189 for the semi-automatic and the classical nucleator, respectively. Taking into account that the time spent analysing one cell is shorter for the semi-automatic nucleator than for the classical nucleator, the semi-automatic nucleator is superior to the classical nucleator.

Shape discrimination by total curvature, with a view to cancer diagnostics.

This paper investigates the use of total curvature for shape discrimination of objects via profiles of their planar sections (not assumed to be star shaped). Methods of estimating total curvature from observation of a finite number of points on the boundary of the object are investigated, including a simple discrete approximation method and various interpolation methods. Total curvature is capable of revealing shape differences on a local scale, as demonstrated by the analysis of two data sets of malignant and normal or benign tumour cell nuclear profiles.

Directional analysis of digitized three-dimensional images by configuration counts.

A method for estimating the orientated rose of normal directions of a three-dimensional (3D) set Z from a digitization of Z, i.e. a voxel image, is presented. It is based on counts of informative configurations in n x n x n voxel cubes. An algorithm for finding all informative configurations is proposed and an estimation procedure is described in detail for the case n= 2. The presented method is a 3D version of a method of estimating the orientated rose of binary planar images using n x n configurations. A new feature is the design-based approach, being more appropriate for biomedical image analysis than the formerly applied model-based approach.

Directional analysis of digitized planar sets by configuration counts.

The oriented rose of normal directions of a random set Z can be used to quantify its anisotropy. A method for estimating this quantity from a digitization of Z in a sampling window, i.e. its pixel image, is presented. The image is analysed locally by considering pixel squares of size n x n. This allows us to count the number of different types of n x n configurations in the pixel image. We show that it suffices to restrict attention to the so-called informative configurations. The number of informative configurations increases only polynomially in n. An algorithm to find these informative configurations is presented. Furthermore, estimators of the oriented rose based on counts of observed informative configurations are derived. The procedure is illustrated by a simulated example and an analysis of a microscopic image of steel.

On the orientational analysis of planar fibre systems from digital images.

The orientational characteristics of fibres in digital images are studied. The fibres are modelled by a planar Boolean model whose typical grain is a thick (coloured) fibre. The aim is to make stereological inference on the rose of directions of the unobservable central fibres from observations made on a digital image of the thick fibres. For central fibres, the relation between the rose of directions and the point intensity, observed on a sampling line, is known. We derive, under regularity conditions, the relation between the unobservable point intensity and the scaled variogram observed on the line in a binary and a greyscale image. Using such a relation, it is possible to draw inference about the rose of directions from the scaled variogram, which is easy and quick to determine in a digital image.