Motivation of a new approach for shape reconstruction based on FBG-optical fibers: Considering of the Bragg-gratings composition as a sensornetwork

In various fields of application, the shape and the tip position of flexible, snakelike objects have to be reconstructed. For this, the considered objects are fitted with so-called shape sensors. This shape sensors are e.g. applied in medical technology to support minimally invasive surgical interventions by tracking flexible instruments; this way navigation systems can be considerably supported. The sensors consist of a solid snakelike body out of flexible carrier material, as silicone, with embedded FBG - optical glass fibers along the object-axis. Guided along the observed instruments, the sensor is supposed to detect the instruments shape by detecting its own ones. The fibers measure the strain at discrete points along the sensor body, which is caused by deformation of the sensor. From these values the shape is estimated. This estimation is performed using specific algorithms. Accordingly, certain requirements regarding the position, orientation and exact number of the measurement units are made. As part of the manufacturing process of the sensor, however, exact control of fiber positioning cannot be realized. To compensate this inaccuracy and also further occurring problems, a fundamentally new calculation approach is presented in this paper. The basic idea is, to consider the system of measurement units as a sensor network. The position and orientation of the units are not considered to be static, because they can only be detected after production but cannot be exactly implemented in a controlled way with a planned position and orientation. The idea is realized by initializing a tensor field on a manifold, representing the surface of the object. This allows to apply the algorithm to measurement values, measured at randomly distributed positions along the sensor body. The new approach is promising and more accuracy in shape sensing is expected do be achieved. The approach of surface characterization is developed in a way that it is transferable to other applications. In the future, also areas in general can be analysed by applying to adapted algorithms based on the same idea. Interpolation of e.g. temperature- and radiation fields can be done in an intelligent way by measuring discrete values by efficiently distributed measurement units.