Varduhi Yeghiazaryan

Automated 3D renal segmentation based on image partitioning

Despite several decades of research into segmentation techniques, automated medical image segmentation is barely usable in a clinical context, and still at vast user time expense. This paper illustrates unsupervised organ segmentation through the use of a novel automated labelling approximation algorithm followed by a hypersurface front propagation method. The approximation stage relies on a pre-computed image partition forest obtained directly from CT scan data. We have implemented all procedures to operate directly on 3D volumes, rather than slice-by-slice, because our algorithms are dimensionality-independent. The results picture segmentations which identify kidneys, but can easily be extrapolated to other body parts. Quantitative analysis of our automated segmentation compared against hand-segmented gold standards indicates an average Dice similarity coefficient of 90%. Results were obtained over volumes of CT data with 9 kidneys, computing both volume-based similarity measures (such as the Dice and Jaccard coefficients, true positive volume fraction) and size-based measures (such as the relative volume difference). The analysis considered both healthy and diseased kidneys, although extreme pathological cases were excluded from the overall count. Such cases are difficult to segment both manually and automatically due to the large amplitude of Hounsfield unit distribution in the scan, and the wide spread of the tumorous tissue inside the abdomen. In the case of kidneys that have maintained their shape, the similarity range lies around the values obtained for inter-operator variability. Whilst the procedure is fully automated, our tools also provide a light level of manual editing.

A family of boundary overlap measures for the evaluation of medical image segmentation

Abstract. All medical image segmentation algorithms need to be validated and compared, yet no evaluation framework is widely accepted within the imaging community. None of the evaluation metrics that are popular in the literature are consistent in the way they rank segmentation results: they tend to be sensitive to one or another type of segmentation error (size, location, and shape) but no single metric covers all error types. We introduce a family of metrics, with hybrid characteristics. These metrics quantify the similarity or difference of segmented regions by considering their average overlap in fixed-size neighborhoods of points on the boundaries of those regions. Our metrics are more sensitive to combinations of segmentation error types than other metrics in the existing literature. We compare the metric performance on collections of segmentation results sourced from carefully compiled two-dimensional synthetic data and three-dimensional medical images. We show that our metrics: (1) penalize errors successfully, especially those around region boundaries; (2) give a low similarity score when existing metrics disagree, thus avoiding overly inflated scores; and (3) score segmentation results over a wider range of values. We analyze a representative metric from this family and the effect of its free parameter on error sensitivity and running time.

Boundary overlap for medical image segmentation evaluation

All medical image segmentation algorithms need to be validated and compared, and yet no evaluation framework is widely accepted within the imaging community. Collections of segmentation results often need to be compared and ranked by their effectiveness. Evaluation measures which are popular in the literature are based on region overlap or boundary distance. None of these are consistent in the way they rank segmentation results: they tend to be sensitive to one or another type of segmentation error (size, location, shape) but no single measure covers all error types. We introduce a new family of measures, with hybrid characteristics. These measures quantify similarity/difference of segmented regions by considering their overlap around the region boundaries. This family is more sensitive than other measures in the literature to combinations of segmentation error types. We compare measure performance on collections of segmentation results sourced from carefully compiled 2D synthetic data, and also on 3D medical image volumes. We show that our new measure: (1) penalises errors successfully, especially those around region boundaries; (2) gives a low similarity score when existing measures disagree, thus avoiding overly inflated scores; and (3) scores segmentation results over a wider range of values. We consider a representative measure from this family and the effect of its only free parameter on error sensitivity, typical value range, and running time.