Toshiya Nakaguchi1, Harufumi Makino2, Tatsuo Igarashi2,
Kenji Kamimura1, Norimichi Tsumura1, Yoichi Miyake1,2
1Dept. of Information and Image Sciences, Chiba University
2Research Center for Frontier Medical Engineering, Chiba University
Transacsion of the Japanese Society for Medical and Biological Engineering, Vol.43, No.4, pp.685-693(2005).
Abstract
In the conventional laparoscopic surgery, one medical doctor who only holds the
laparoscope during the surgery must be needed. Therefore in this study, we
implement the automatic tracking and zooming system for laparoscopic surgery
by means of a steady laparoscope on the abdominal surface. The system
can be additionally installed to the conventional laparoscope system and
it analyses camera images in real-time. In this system the markers attached
at the top of the forceps are extracted, and the focusing area can be estimated.
In this study, we implement two ways of marker tracking algorithms; one
tracks a single marker on the forceps which is assumed as focusing point,
and another tracks two different colored markers, which can estimate the
more practical focusing point. In order to perform high-quality real-time
movie processing, we implement the system by using CPU and GPU (Graphics
Processing Unit) in parallel. The tracking algorithm is carried out by
CPU and the image enlargement is carried out by GPU, then we obtain high
frame rate and low delay time. For the automatic smooth camera work, we
propose a new control model which has many parameters. Through six times
of animal experimentation, we obtained the appropriate values of parameters
evaluated by two experienced doctors. After the evaluation of the experienced
doctors, we can obtain the effectiveness of the system, and they can operate
the cholecystectomy successfully by the animal experiments without supporting
by any scope holding doctor.
Text and Figures ( to appeare )
[PDF] [PDF figures](in Japanese)