![](bg5.png)
(a) (b)
Figure 6: Evaluation of the screening quality: left - time
spent in sectors - lighter color means longer time; right -
density of coverage in sectors - lighter color means more ho-
mogeneous examination with less holes (neglected regions).
We demonstrate the visualization output of the US
setup on two real examples. The first one is a screen-
ing examination, during which the whole breast and
adjacent lymph nodes should be covered. The sec-
ond one is a diagnostic examination, during which
the physician focuses only on areas with earlier find-
ings. The transducer trajectory for both examinations
has been shown in Fig. 4. The trajectory is rendered
in the reference coordinate system R and the time is
coded in color. In the screening example Fig. 4(a),
the physician started in the axillary lymph nodes and
continued with a systematic examination of the breast
from the areola to its sides. In the diagnostic example
Fig. 4(b), only a small section of the breast is covered.
The differences between both examinations are even
more noticeable in our second type of visualization in
Fig. 5, which shows the time spent in each voxel. The
voxel color denotes the duration in seconds, for which
the corresponding area has been examined. We use
semitransparency of voxels to better visualize the vol-
ume. Note that in the screening example Fig. 4(a) the
most frequently visualized area is the areola, whereas
in the diagnostic example it is the area of the earlier
finding, which in this case is in the upper outer quad-
rant. In addition, we can perform 3D reconstruction
of the area and apply segmentation algorithms to 3D
reconstruction, such as GrowCuts (Zhu et al., 2014),
and visualize present lesions as shown in Fig. 7.
4 CONCLUSION
We have proposed an affordable enhancement of US
devices, which is currently being clinically tested. An
electro-magnetic tracking technology is used to si-
multaneously track the position of a patient and a free-
hand US transducer, which allows us to perform sev-
eral visualization tasks. We can for example display
spatial distribution combined with duration, for which
every location in our 3D space has been examined.
Figure 7: Example of 3D segmentation using GrowCuts.
The advantages of the proposed setup are twofold.
We can perform blind tests and evaluate physicians
during US examinations or we can guide the physi-
cians to regions insufficiently examined and show 3D
reconstruction of lesions.
Possible alternative to using the electro-magnetic
tracker would be to use an optical tracking system,
such as the Microsoft Kinect. In our opinion, how-
ever, the E-M tracker provides more robust opera-
tion with more accurate data, and, with fewer require-
ments on positioning, direct visibility, etc., it imposes
less stress both on the expert performing the examina-
tion, and on patient. Furthermore, using some video
tracking system, additional ethical issues could possi-
bly arise.
Future work includes an evaluation study of exam-
ination procedures over a larger number of specialists
and patients.
ACKNOWLEDGMENTS
This work has been supported by the Technol-
ogy Agency of the Czech Republic (TACR) Project
TA04011392.
REFERENCES
Andrei, S. S., Gao, J., Chiu, J., Turian, J., and Caluser, C.
(2014). Performance evaluation of an automated free-
hand breast ultrasound system for the completeness of
scanning assessment. European Congress of Radiol-
ogy, C-1382.
Boyd, N. F., Guo, H., Martin, L. J., Sun, L., Stone, J.,
Fishell, E., Jong, R. A., Hislop, G., Chiarelli, A.,
Minkin, S., and Yaffe, M. J. (2007). Mammographic
density and the risk and detection of breast cancer.
New England Journal of Medicine, 356(3):227–236.
BIOIMAGING 2017 - 4th International Conference on Bioimaging
132