|ISSN No. 1606-7754 Vol.13 No.1 April 2005|
First-order stereology in diabetes and endocrine research-number and volume estimation of objects
Ibrahim M Inuwa
Department of Human and Clinical Anatomy, College of Medicine and Health Sciences, Sultan Qaboos University, Muscat, Sultanate of Oman
Stereology is a technique that enables acquisition of data on number, volume, length or surface area of identifiable objects in a three-dimensional structure by sampling in two dimensions. That is, it provides a technique for quantifying objects on a slice from the structure such as a histological specimen viewed under the microscope or a computerised tomography (CT) scan. It has the enormous virtue of having a rigorous mathematical foundation and rules for counting that also give a reliable measure as well as an indication of precision. First-order stereology denotes estimating volume, surface area, length or number of any biological object. Terms such as degeneration, toxicity, atrophy hypertrophy, dysgenesis, and proliferation all refer to alterations in one or more of these parameters. However, as it is the case with almost all pertinent studies, it is not possible to count the number of relevant units directly. Yet the information is essential if we want to know how experimental or environmental interventions affect an organ’s development. For example, what effect does the absence of a particular growth factor or reduced nutrition have on the number or size of islets, and at what stage of development does this occur; or has behavioural experience or environmental insult affected the number of chromophobes in the pituitary gland? The mainly unbiased and reliable way to obtain such information is by means of stereology. This technique is an absolutely essential tool for any biologist who needs to know the number of units in any system — whether they be functional units like number of islets within the pancreas, or cell organelles like mitochondria. It also provides a means of obtaining size, length and surface area of any object under consideration. In this review the current practical applications of two first-order stereology parameters in relation to diabetes and endocrine research are highlighted.
Keywords: Diabetes, Measurement, Stereology
The term stereology, which has its origin in the Greek word stereos (or solid), was introduced into the scientific vernacular in the early 1960s.1-2 It is a term that describe a discipline focused on analyzing three dimensional (3-D) structural parameters of objects based on their appearance on two dimensional (2-D) histological sections.3-5
A pathologist making a judgement of cause of death or a metallurgist making an assessment of the cause of machinery failure are examples of situations where a scientist makes a qualitative analysis. In these and similar circumstances that is all that may be required to answer the underlying scientific question. However, in many other fields of applied science a scientist is expected to provide a quantitative analysis. This brings with it the responsibility to deal seriously with issues of accuracy, precision and overall confidence in the results of an investigation. In other words, a more rigorous approach must be adopted. Stereology attempts to provide quantitative data6 by ensuring a rigorous sampling regime, coupled with robust mathematically sound estimators.7-10 With the initiation by Delesse in 1847,11 but mainly since the beginning of the twentieth century, many stereological methods have been published allowing us to relate two-dimensional measurements easily obtainable on flat histological images with three-dimensional characteristics of the structure analysed.
Primacy of sampling
Stereology is fundamentally statistical in its nature, the methods rely upon careful sampling design and a robust sampling theory.12 The methods cannot properly be applied unless a random sample of one form or another has been taken. Random sampling13-16 must be applied in subjects (human and animal models) selection and at all the sampling levels required within an individual subject, for example, tissue blocks, histological sections, microscope fields of view and individual measurements. The procedure for applying randomness is achieved either independently (eg using random numbers table) – where all subjects in the population have independent equal probability of being selected, or systematically10 where the first object is selected randomly and then each successive object is sampled systematically (Fig 1).