=Paper=
{{Paper
|id=Vol-1475/paper34
|storemode=property
|title=Ultrasound Segmentation in Navigated Liver Surgery
|pdfUrl=https://ceur-ws.org/Vol-1475/Proceedings_CURAC_2010_Paper_34.pdf
|volume=Vol-1475
|dblpUrl=https://dblp.org/rec/conf/curac/AndereggPW10
}}
==Ultrasound Segmentation in Navigated Liver Surgery==
https://ceur-ws.org/Vol-1475/Proceedings_CURAC_2010_Paper_34.pdf
Ultrasound Segmentation in Navigated Liver Surgery
Sylvain Anderegg, Matthias Peterhans, Stefan Weber
University of Bern, Institute for Surgical Technology & Biomechanics (ISTB), Bern, Switzerland
Contact: matthias.peterhans@istb.unibe.ch
Abstract:
In computer assisted liver surgery, the use of pre-operative 3D computer tomography (CT) images provides basic orien-
tation and valuable information about the patient liver. During the surgical intervention, the accuracy of this informa-
tion is reduced by motion and deformation of the organ. The fusion of intra-operative ultrasound (US) imaging with the
pre-operative data is the next step in order to improve this situation. As pre-requisite, the identification of correspond-
ing structures in US and CT (such as blood vessels) is required. Within this paper, the integration of an ultrasound ves-
sel segmentation algorithm in a navigation system for liver surgery is presented. Initial results obtained on patients un-
dergoing liver resection are evaluated.
Keywords: Liver surgery, Ultrasound segmentation, Computer assisted surgery
1 Problem
Intra-operative ultrasound imaging is used during surgical interventions on the open liver, providing real-time informa-
tion about the organ inner structure. Moreover, new tumors may be detected and the surgical intervention shall eventual-
ly be adapted on site. Within our navigation system for liver surgery [1], [2], surgeons are provided with the pre-
operative 3D images, navigated US imaging and a set of optical trackers for the navigation of standard surgical tools.
During the surgical intervention, the accuracy of the pre-operative images is reduced due to motions and deformations
of the liver. Thus, updating the pre-operative images with the 3D information obtained from US imaging is an important
step towards an increase of the navigation system accuracy.
Registration of CT images with US imaging is usually performed by detecting corresponding features in both data
types. A non-rigid transformation can then be computed and applied onto the CT images like in [3]. The most obvious
common features between CT- and US- images are liver vessels. They are easily segmented pre-operatively in the CT
images, whereas the segmentation of vessels in US images is more challenging as these images have a low signal-noise
ratio. Furthermore, it is strongly desirable that the segmentation happens in real-time in order to provide surgeons with
direct feedback on the quality of the acquired US information. Several groups have reported on US segmentation algo-
rithms in liver surgery [4] but most of them focused on tumor detection. Recently, [5] showed an algorithm for vessel
segmentation and tracking, which requires an initial manual segmentation to start. A dynamic texture approach was pre-
sented in [6]. Good segmentation results are reported but again, user intervention is required for tissue classification. In
[7] a vessel segmentation algorithm was developed and tested on several porcine livers under isolated perfusion. This
segmentation algorithm gave promising results both in terms of efficiency and computational time and does not require
user interaction. Within this paper we present the integration of this segmentation algorithm in our navigation system
and the initial results of its use in real conditions.
2 Methods
A. Navigation System and Ultrasound Device
The surgical navigation system is built as a transportable setup containing an NDI Vicra Camera (Northern Digital Inc,
Canada), an integrated Terason T3000 ultrasound system with an 8IOA intra-operative probe (Teratech Corporation,
USA) and a Shuttle barebone PC (Shuttle Inc, Taiwan) with a touch-screen monitor (ELO Touchsystems, USA). In-
strument tracking is enabled by a navigation toolset composed of adapters to the existing surgical tools (CUSA, ablation
devices, US probe) and a pointer calibration unit.
Proceedings curac2010@MEDICA 173
B. Workflow
Before each navigated liver surgery, pre-operative tri-phase CT data was processed by MeVis Distant Services. The
MeVis analysis provides segmentation of all the important structures (vessels, tumors, surface) as well as several resec-
tion proposals. The resulting 3D models are evaluated by the surgeons and appropriate visualization models are selected
and loaded into the navigation system.
The navigated surgery starts by an initial rigid registration of the liver. Four landmarks are manually defined on the vir-
tual liver model. Then, the tip of the CUSA is calibrated with the pointer calibration unit. Using the CUSA tip, the
surgeon points the four manually defined landmarks on the real liver, which are automatically recorded by the naviga-
tion system. Through the alignment of these two point sets, initial rigid registration is obtained.
In a second stage, US images are acquired and segmented in real-time. The US probe is pre-calibrated using an US cali-
bration unit pre-operatively [8]. Using the probe, vessels are traced by the surgeon. The obtained segmentation and im-
ages are recorded with the corresponding position tracking information for further data analysis.
C. Vessel Segmentation
The vessel segmentation algorithm is based on the approach proposed by Dagon in [7] may be shortly described as:
1. Vessel image mask generation
2. Vessel seed point detection
3. Vessel approximation using an elliptical model
4. Outliers removal
The vessel mask generation starts with a scale-space blob detection [9], which is performed by generating scale-space
images Il with eight different scale levels Sl using a Gaussian Kernel of:
σ = 3 ⋅ Sl (1)
The pixel intensities of the scale-space images are then summed in a scale-space map IMap with the following equation:
n
I Map ( x, y ) = ∑ [I n ( x, y ) − I l ( x, y )] (2)
l =1
Thus, dark features in the US image that present a large scale-space lifetime are enhanced. A binary threshold tsc is then
applied on the scale-space map to obtain the features binary mask IMask.
The features mask is then cross-correlated with a binary circular mask of radius rc and normalized. A threshold tcorr is
applied on the obtained image and a connected component labeling algorithm is run in order to identify the highest in-
tensity pixel in each region, which is then labeled as seed point.
To approximate the blood vessel, an elliptical model is applied. To detect the vessel border, rays are regularly casted
around the seed point on the initial scale-space mask IMask. The edge points are detected at the first value change in the
ray intensity profile. The edge points are then used to initialize a direct least square ellipse fitting algorithm [10], [11].
Finally, outliers are filtered by defining minimal and maximal ellipse axes ratio and axes average length.
The C++ implementation of the segmentation algorithm was based on the code provided by Dagon [7] which was im-
proved by adding outlier removal in the vessel seed point detection step and by integrating a region of interest (ROI)
provided by the direct interface to an US imaging device (described in Section 2.D)
D. Software integration
The segmentation algorithm was integrated into our navigation software [1],[2]. In this way, the segmentation is applied
in real-time onto the images generated by the US probe. The obtained segmentation is visualized in both the 2D US im-
age and in the 3D world viewer. A touch-screen user interface was developed for easy use in the operation room envi-
ronment. The user can switch the segmentation algorithm on and off, change segmentation parameters and record both
174 Proceedings curac2010@MEDICA
US images and the obtained segmentations. The segmentation steps can be visualized to better understand the parame-
ters respective influences (Fig. 1). The whole sequence can be loaded and segmented again later with different pa-
rameters.
Fig. 1: Navigation software: On the left side, the user interface in detailed mode displays each step of the seg-
mentation algorithm (original image, ROI, scale-space, correlation, seed points, ellipse fitting). On the right
side, the 3D view shows the pre-operative image, tracked US probe and the segmentation (yellow).
E. Validation
A validation application was developed for the result analysis. The recorded US images are loaded again with the ob-
tained automatic vessel segmentation centers. The user then manually defines whether the detected vessel centers are
correct, wrong (false positive) or missing (Fig. 2, right). The application then outputs the total number of each vessel
center classification.
Fig. 2: Typical vessel detection produced by the vessel segmentation algorithm (left). Two segmented vessels
are correct (+) and one is wrong (x) (right).
3 Results
About 2’200 images (300x430 pixels) were recorded and segmented during one surgical intervention at the Inselspital
in Bern, Switzerland. The US probe parameters for brightness, contrast and time gain control were set to the default
values. The average time to process one image was 70ms, which resulted in a frame rate of 7 frames per second. Six
subsets including a total of 538 images were selected for post-operative analysis with optimized segmentation parame-
ters. Table 1 and 2 present the results obtained with the parameters: tsc = 65, rc = 10 pixels, tcorr = 0.01, Region of Inter-
est: 193x371 pixels.
Proceedings curac2010@MEDICA 175
Processed images Expected vessel centers Correct centers Missing centers
Set 0 109 551 250 301
Set 1 42 107 55 52
Set 3 147 463 223 240
Set 6 73 265 73 192
Set 10 13 26 22 4
Set 12 154 408 278 130
Table 1: Amount of correct- and missing- centers compared to the number of expected vessels.
Processed images Vessels detected Correct classification Wrong classification
Set 0 109 374 250 124
Set 1 42 158 55 103
Set 3 147 421 223 198
Set 6 73 228 73 155
Set 10 13 25 22 3
Set 12 154 295 278 17
Table 2: Amount of correct- and wrong- centers compared to the number of detected vessels.
The average value of correctly classified vessels is 60.03%. The average number of missing vessels per frame is 1.7 and
the average value of wrongly classified vessels per frame is 1.1.
4 Discussion
The number of missing vessels per frame is relatively high and is mainly caused by small vessels. When the threshold
value tsc is set to a low value, small vessels are better detected but the amount of wrongly classified vessels from image
noise increases as well. Wrongly classified vessels come from homogeneous noise regions in the images and from arte-
facts lying under the liver borders as illustrated in Fig. 2. When comparing our results with the ones presented by Dagon
[7], we see improved outcomes in the amount of correctly segmented vessels (40% by Dagon, 60% in this work). How-
ever, the results are not directly comparable as different image data and slightly different validation methodologies were
used. The images used herein were acquired in real surgery on human patients where Dagon used an ex-vivo porcine
liver specimen under isolated perfusion. The validation approaches differ in that the distance between the manual and
automatic segmented vessel centers was not considered for the vessel classification within this work. The higher com-
putation time of 70ms vs 25ms in this work can be explained by the larger size of the US images used with the Terason
imaging device and the added load due to the visualization of the segmentation steps as shown in Fig. 1.
As we aim to use the segmentation results for intra-operative registration, a further reduction of wrongly classified ves-
sels is desirable. This should be partly solved by detecting the lower organ borders. Another way for reducing wrongly
classified vessels is to use the 3D position information of the segmented vessels in the prior images to predict their posi-
tions in the next frame. In general, the large number of segmentation parameters made it difficult to obtain a good seg-
mentation while in the operating room. Reducing their number or an automatic adjustment of some parameters might
solve this issue.
Another effect observed when comparing the segmentation results from different acquisitions is the high sensitivity to
the US image quality. We believe that by adjusting the US imaging parameters, in particular brightness, contrast and
time gain control, the number of missing vessels per frame should decrease. The development of the vessel-based regis-
tration methods will show whether a reliable detection of large vessels or detection of smaller vessels with higher error
rate are preferable.
Finally, the computation time of 70ms per frame is not yet acceptable for a real-time application. Considering that there
should be more processing for the registration afterwards. This should be improved by code optimization and/or using
GPU computing or multi-threading.
176 Proceedings curac2010@MEDICA
5 References
[1] M. Peterhans, B. Dagon, L. Nolte, C. Baur, A. Vom Berg, D. Inderbitzin, and S. Weber, “Soft tissue navigation:
Transfering operative planning data into the operation room”, ECR 2009 Book of Abstracts, 2009.
[2] A. Vom Berg, D. Candinas, D. Inderbitzin, M. Peterhans, S. Weber, and L. Nolte, “Computer assisted surgery
and navigation in complex hepatic surgery and tumorablation: first clinical results of 10 patients”, Proceedings
of CARS 2010, 2010.
[3] T. Lange, N. Papenberg, S. Heldmann, J. Modersitzki, B. Fischer, H. Lamecker, and P.M. Schlag, “3D ultra-
sound-CT registration of the liver using combined landmark-intensity information,” IJMRCAS, 2009.
[4] J. A. Noble, and D. Boukerroui, “Ultrasound Image Segmentation: a Survey”, TMI, 2006.
[5] J. Guerrero, S. E. Salcudean, J. A. McEwen, B. A. Masri, and S. Nicolaou, “Real-Time Vessel Segmentation
and Tracking for Ultrasound Imaging Applications”, TMI, 2007.
[6] S. Milko, E. Samset, and T. Kadir, “Segmentation of the liver in ultrasound: a dynamic texture approach”,
IJMRCAS, 2008.
[7] B. Dagon, C. Baur, and V. Bettschart, “Real-Time Update of 3D Deformable Models for Computed Aided Liv-
er Surgery”, Proceedings of IAPR, 2008.
[8] M. Peterhans, S. Anderegg, P. Gaillard, T. Oliveira-Santos, and S. Weber, “A Fully Automatic Calibration
Framework for Navigated Ultrasound Imaging”, Accepted for Proceedings of EMBC 2010, 2010.
[9] T. Lindeberg, “Detecting Salient Blob-Like Image Structures and Their Scales with a Scale-Space Primal
Sketch: A Method for Focus-of-Attention”, International Journal of Computer Vision, 1993.
[10] A. Fitzgibbon, M. Pilu, and R. B. Fisher, “Direct Least Square Fitting of Ellipses”, 13th International Confe-
rence on Pattern Recognition, 1996.
[11] R. Halir and J. Flusser, “Numerically Stable Direct Least Squares Fitting of Ellipses”, 1998.
Proceedings curac2010@MEDICA 177