Abstract
Quantification of cell populations in tissue sections is frequently examined in studies of human disease. However, traditional manual imaging of sections stained with immunohistochemistry is laborious, time-consuming, and often assesses fields of view rather than the whole tissue section. The analysis is usually manual or utilises expensive proprietary image analysis platforms. Whole-slide imaging allows rapid automated visualisation of entire tissue sections. This approach increases the quantum of data generated per slide, decreases user time compared to manual microscopy, and reduces selection bias. However, such large data sets mean that manual image analysis is no longer practicable, requiring an automated process. In the case of diabetes, the contribution of various pancreatic endocrine cell populations is often investigated in preclinical and clinical samples. We developed a two-part method to measure pancreatic endocrine cell mass, firstly describing imaging using an automated slide-scanner, and secondly, the analysis of the resulting large image data sets using the open-source software, Fiji, which is freely available to all researchers and has cross-platform compatibility. This protocol is highly versatile and may be applied either in full or in part to analysis of IHC images created using other imaging platforms and/or the analysis of other tissues and cell markers.