This workshop is partially supported
by Australian Research Council,
discovery project DP140102794 &
ARC Future Fellowship (FT110100623).
ARC Logo 


The dataset consists of a collection of 17 multi-layer cervical cell volumes, from which 8 will be used for training and 9 for testing. Details of the dataset will be released later.

Dataset Release Schedule
  • December 1st, 2014: Release of 8 annotated multi layered cytology volumes, where four are part of the training set and four are from the testing set. The annotation contains the delineation of each cervical cell (cytoplasm and nucleus). We will also release the evaluation code written in Matlab.

  • January 1st, 2015: Release of the remaining 9 multi layered cytology volumes. We will provide manual annotation only for the training volumes.


  • Annotation for Testset

    • Annotation for Test set

      • Please cite the dataset by the following paper “Zhi Lu, Gustavo Carneiro, and Andrew P. Bradley. An Improved Joint Optimization of Multiple Level Set Functions for the Segmentation of Overlapping Cervical Cells. IEEE Transactions on Image Processing. Vol.24, No.4, pp.1261-1272, April 2015.”

How to Use Evaluation Function?

The evaluateion code is used to quantitatively assess the segmentation performance. It consists of two functions:

  • evaluateCytoSegmentation.m: Function to evaluate the result.

  • SegEvaluateJIDiceTPRFPR.m: Function about the evaluation measurement, called by evaluateCytoSegmentation.m.

You need to load the cytoplasm annotation of each cell into a Matlab Cell structure before you call the function evaluateCytoSegmentation.m to start evaluating. For example, cell GroundTruth{numImage_i,1}{CellID_j,1} refers to the j-th cell annotation of image i. Therefore,we suggest your results of segmentation are organized in the same way.