Minimally invasive surgical interventions performed with the assistance of computerized navigation systems require reliable registration methods for preoperatively acquired patient anatomy representations. These registration methods have to be compatible with the minimally invasive paradigm. The use of the non-invasive brightness-mode ultrasound, which is known not to be harmful to the patient, seems to be especially promising for that purpose. However, associated devices should preferably work in a computationally efficient and fully automatic manner in order to minimize user interactions and in particular to avoid manual segmentation tasks. This paper presents a fast and fully automatic segmentation approach for ultrasound B-Mode images, which is capable of detecting echoes originating from bony structures. As opposed to algorithms that are designed for surface reconstructions which need the detected contours to be complete in the ultrasound image, our solution focuses more on the precise and rapid detection of bone contours that could potentially be used for a minimally invasive registration procedure. In addition an overview of the developed image processing scheme, an experimental setup is described that enables a direct comparison between manually digitized reference points and the segmented bone contour. The practically achievable segmentation accuracy of the proposed algorithm is presented using a cadaveric study based on bovine and porcine legs. The results show a good concordance of the automatically generated contours in respect to the real position of the bony surface. IEEE TRANSACTIONS ON MEDICAL IMAGING 3
[1]
Kristel Michielsen,et al.
Morphological image analysis
,
2000
.
[2]
T. Laine,et al.
Accuracy of pedicle screw insertion with and without computer assistance: a randomised controlled clinical study in 100 consecutive patients
,
2000,
European Spine Journal.
[3]
David J. Hawkes,et al.
AcouStick: A Tracked A-Mode Ultrasonography System for Registration in Image-Guided Surgery
,
1999,
MICCAI.
[4]
Jacques Demongeot,et al.
Automated Registration of Ultrasound with CT Images: Application to Computer Assisted Prostate Radiotherapy and Orthopedics
,
1999,
MICCAI.
[5]
R A Peters,et al.
Automatic segmentation of ultrasound images using morphological operators.
,
1991,
IEEE transactions on medical imaging.