Multi-Organ Segmentation of CT Images using Statistical Region Merging

Abstract

Segmentation is one of the key steps in the process of developing anatomical models for calculation of safe medical dose of radiation for children. This study explores the potential of the Statistical Region Merging segmentation technique for tissue segmentation in CT images. An analytical criterion allowing for an automatic tuning of the method is developed. The experiments are performed using a data set of 54 images from one patient, demonstrating the validity of the proposed criterion. The results are evaluated using the Jaccard index and a measure of border error with tolerance which addresses, application-dependant, acceptable error. The outcome shows that the technique has a great potential to become a method of choice for segmentation of CT images with an overall average boundary precison, for six representative tissues, equal to 0.937. KEY WORDSVoxel model, image segmentation, statistical region merging.

Full-Body CT Segmentation using 3D Extension of Two Graph-based Methods: A Feasibility Study

“…Both methods were successful in segmenting complex medical images in several reported studies (e.g. [17], [18], [19], [20]). …”

“…Thus, it is desirable to consider the smallest Q value sufficient for region separation. This can be fairly well estimated in 2D setting with an analytical criterion ( [20], [18]) but not in 3D contex due to the different image intensity characteristics in larger number of adjacent tissues, thus, an undermerging is unavoidable.…”

“…Both methods are fast enough to process CT patient data in real time and each requires only one parameter to be set up. However, given three-dimensional volumetric data such as CT scans, a direct segmentation in 3D voxel set is more desirable than 2D (slice by slice) segmentation to overcome overmerging issues reported in [20]. The 3D approach we propose in this study allows for a simultaneous tracking of boundaries of elongated objects allowing information about boundaries to pass freely between slices.…”

3D Segmentation for Multi-Organs in CT Images

“…Successful medical image segmentations using these methods were reported in several studies (e.g. [18], [19], [20], [21]), [22]), despite the fact that the methods assume some homogeneity property for components, which often fails in medical images.…”

“…In some specific applications like masses segmentation in mammograms ( [19]) or segmentation of individual CT slice images ( [21]), it is possible to develop an analytical criterion helping in optimizing the Q value. That is, one can estimate the smallest value of Q sufficient for successful segmentation of regions of interests e.g.…”