Novel context-based segmentation algorithms for intelligent microscopy.


You have reached the Intelligent Microscopy Project website. This project is currently funded by the Engineering and Physical Sciences Research Council of Great Britain (EPSRC) grant number EP/M023869/1. Our aim is to develop new intelligent computer methods to analyse, quantify and understand the information contained in images of cells and tissues obtained using digital microscopy.

Intelligent Microscopy

By 'intelligent' we refer to the explicit development of algorithmic procedures on abstracted spatial information that allows some degree of mechanical reasoning about its interpreted content. In our case the spatial information is extracted from digitised images of histological preparations (e.g. stained tissue sections) examined under the microscope.

Currently, the assessment of some features used in histopathological diagnosis, rely to a great extent on expert observers (histopatholgists) and their experience. In some aspects, this has the disadvantage that an inevitable element of subjectivity in visual observation makes it difficult to reach quantitative or reproducible (inter- and also intra-observer-wise) judgements about certain histological features.

Project aims

We are designing 'context-based' imaging programs to help advance automated analysis and histopathological cancer diagnosis. By 'context-based' here we understand firstly, the use of data constructs that allow the structure and relations of cells and tissues in histological samples to be modelled in a way that enables computer programs to subsequently 'reason' mechanically about the image contents, and secondly, to allow methods of data extraction that are both quantitative and reproducible, but also represent the biological realities where the data is extracted from.

The potential for intelligent algorithms in histological imaging have wide applications in various areas of biomedicine. This is especially true in histopathology, where in relation to cancer diagnosis, it could enable to more accurately modelling and prediction of the status of malignant disease to target and implement the most suitable forms of treatment.

We are investigating analytical tasks using a spatial logic called Discrete Mereotopology (DM) and have produced proof-of-concept software and peer-reviewed publications which demonstrate the efficacy of this approach for encoding and querying relations between the biologically-relevant entities of interest (e.g. cell components, tissue layers, staining patterns) and the image segments detected to correspond to them. This enables the formulation of histologically relevant models (e.g. cells, tissue types and voids) that can be operated on at a level that has not been possible before. This is a departure from traditional pixel-based routines, leading to region-based algorithms that are histologically relevant to the range of images and structures that are expected in histological samples.

This logic-based approach to image analysis brings several advantages. Firstly, it provides a robust, rigorous mathematical foundation for software development. Secondly, it allows histological images to be systematically interpreted in terms of histologically-relevant entities, not just pixels. Thirdly, it enables symbolic and automated reasoning programs to be used alongside numerical methods.

We are also aiming to develop markers of data quality to identify and correct some microscopy artefacts as well as label structures according to how well they approach the expected models they represent. In this way indications of 'confidence in the results' can be associated to the models found in microscopy images.

We believe that the approach outlined here is both translational and invaluable in most biological areas using microscopy, where quantitative results are required to make sound evidence-based decisions.

This is a multidisciplinary project covering image analysis, computer science and pathology, and whose members are:

Interested in joining the group?

Please let us know, we are interested in a number of subjects that can form the basis for a doctoral research programme.


  1. Randell DA, Landini G. Discrete mereotopology in automated histological image analysis. Proceedings of the Second ImageJ user and developer Conference, Luxembourg, 6-7 November, 2008, p 151-156. ISBN 2-919941-06-2
  2. Landini G, Randell DA, Galton. Intelligent imaging using Discrete Mereotopology. Proceedings of the Fourth ImageJ user and developer Conference, Luxembourg, 24-26 October, 2012, p 99-103. ISBN:2-919941-18-6.
  3. Landini G, Randell DA, Galton.Discrete Mereotopology in Histological Imaging. In: Ela Claridge, Andrew D. Palmer and William T. E. Pitkeathly (editors), Medical Image Understanding and Analysis, Proceedings of the 17th Conference on Medical Image Understanding and Analysis, 17th-19th July 2013, Birmingham, UK, pages 101-106. PDF
  4. Sioutis, M.; Condotta, J-F.; Salhi, Y.; Mazure, B.; Randell, D.A. (2015): 'On Ordering SpatioTemporal Sequences to meet Transition Constraints: Complexity and Framework', In: Proc AIAI-2015: 130-150.
  5. Flight R. Landini G, Styles I, Shelton R, Milward M, Cooper P. Automated optimisation of cell segmentation parameters in phase contrast microscopy using discrete mereotopology. Proceedings of the 19th Conference on Medical Image Understanding and Analysis. Lincoln, Jul 15-17, 2015. PDF
  6. Randell DA, Landini, G, Galton A. Discrete Mereotopology for spatial reasoning in automated histological image analysis. IEEE Trans Patt Rec Mach Intell 35(3): 568-581, 2013. DOI: 10.1109/TPAMI.2012.128 PDF
  7. Galton A, Landini G, Randell D, Fouad S. Ontological levels in histological imaging. 9th International Conference on Formal Ontology in Information Systems, FOIS 2016, Annecy, Jul 4-6, 2016. PDF
  8. Fouad S, Landini G, Randell D, Galton A. Morphological separation of clustered nuclei in histological images. ICIAR 2016, Lecture Notes in Computer Science, 9730, pp. 599–607, 2016. Presented at 13th International Conference on Image Analysis and Recognition, ICIAR 2016, Póvoa de Varzim, Portugal, July 13-15, 2016. PDF
  9. Landini G, Randell D, Fouad S, Galton A. Automatic thresholding from the gradients of region boundaries. Journal of Microscopy 2016 (in press). PDF
  10. Florindo JB, Landini G, Bruno OM. Three-dimensional connectivity index for texture recognition. Pattern Review Letters 2016 (in press). PDF


We have chosen to develop our imaging algorithms in ImageJ (multiplatform and open source) for ease of development and to assure their impact, so they are available without complicated or restricting licenses or specialised hardware. We only request that if you use our routines in your work, please acknowledge the associated publications, and also let us know of your work.

Here is our preliminary implementation of Discrete Mereotopology in 2D for region connection calculus we use for spatial reasoning.

More ImageJ software