Segmentation of liver, its vessels and lesions from CT images for surgical planning

Abstract

Background: Cancer treatments are complex and involve different actions, which include many times a surgical procedure. Medical imaging provides important information for surgical planning, and it usually demands a proper segmentation, i.e., the identification of meaningful objects, such as organs and lesions. This study proposes a methodology to segment the liver, its vessels and nodules from computer tomography images for surgical planning.

Methods: The proposed methodology consists of four steps executed sequentially: segmentation of liver, segmentation of vessels and nodules, identification of hepatic and portal veins, and segmentation of Couinaud anatomical segments. Firstly, the liver is segmented by a method based on a deformable model implemented through level sets, of which parameters are adjusted by using a supervised optimization procedure. Secondly, a mixture model is used to segment nodules and vessels through a region growing process. Then, the identification of hepatic and portal veins is performed using liver anatomical knowledge and a vein tracking algorithm. Finally, the Couinaud anatomical segments are identified according to the anatomical liver model proposed by Couinaud.

Results: Experiments were conducted using data and metrics brought from the liver segmentation competition held in the Sliver07 conference. A subset of five exams was used for estimation of segmentation parameter values, while 15 exams were used for evaluation. The method attained a good performance in 17 of the 20 exams, being ranked as the 6th best semi-automatic method when comparing to the methods described on the Sliver07 website (2008). It attained visual consistent results for nodules and veins segmentation, and we compiled the results, showing the best, worst, and mean results for all dataset.

Conclusions: The method for liver segmentation performed well, according to the results of the numerical evaluation implemented, and the segmentation of liver internal structures were consistent with the anatomy of the liver, as confirmed by a specialist. The analysis provided evidences that the method to segment the liver may be applied to segment other organs, especially to those whose distribution of voxel intensities is nearly Gaussian shaped.

Figure 1

Liver Functional Anatomy: segments of Couinaud. Hepatic veins and portal vein define the segments, except segment I, which is drained by cava vein.

Figure 2

P(x) definition. The speed image is defined by a function based on two automatically defined thresholds T_L and T_H (T_L<T_H) and the input image g(x)

Figure 3

Gaussian thresholds estimation. The two thresholds T_L and T_H are defined as the values for which the Gaussian returns two pre-defined values, respectively G_L and G_H

Figure 4

Fitness evaluation. The process evaluates a given set of parameters comparing reference data and the result generated by segmentation using the given parameters

Figure 5

Gaussian mixture model. The figure shows the mixture model assumed to segment the liver parenchyma (center gaussian), vessels (right gaussian) and nodules (left gaussian).

Figure 6

Thresholds definition. The thresholds T_L^lc and T_H^rc are set empirically as the intensity value in which the proportion between voxels belonging to the given Gaussian and voxels belonging to the mixture is at least 70%. The thresholds T_L^rc and T_H^lc are defined as the intensity value where the Gaussians intercept.

Figure 7

hepatic vein main branches identification. The hepatic branches are identified clock wisely: right, medium, and left branches.

Figure 8

Best result obtained: axial view; coronal view; sagittal view. The liver segmentation result appears in green.

Figure 9

Leak caused by a big nodule: axial view; coronal view; sagittal view. The liver segmentation result appears in green.

Figure 10

Analysis of parameters sensibility: RMS, G_L, G_H, propagation weight, and mean curvature weight. The charts show the range of values and their performance result for each parameter.

Figure 11

Vessels and nodules segmentation: axial, coronal, sagittal and 3D views. The liver appears in light green, the vessels in red and the nodules in dark green.

Figure 12

Hepatic vein main branches identification. The right main branch appears in green, the medium in yellow, and the left in blue. The other vessels appear in red.

Figure 13

Hepatic and Portal veins segmentation. The hepatic veins appear in red and the portal vein in blue.

Figure 14

Couinaud segmentation: axial, coronal, sagittal and 3d views. The Couinaud segments appear in different colors, following the model described in figure 1.