Major aspects on long-term outcome in Total Knee Arthroplasty are the correct alignment of the implant with the mechanical load axis, the rotational alignment of the components as well as good soft tissue balancing. To reduce the variability of implant alignment and at the same time minimise the invasiveness different computer assisted systems have been introduced.

To achieve accuracy as high as those of a robotic system but with a pure mechanically adjustable cutting block, the Exactech GPS system has been developed. The new concept comprises a seamlessly planning and navigation screen with an integrated optical tracking system for fast and accurate acquisition and verification of anatomical landmarks within the sterile field as well as a tiny cutting guide for accurate transfer of the planned bone resections.

Using a conventional screwdriver the cutting block could be accurately aligned with the planned resection by controlling the current position of the cutting block on the navigation screen. To save time, to maximise the ease of use and to minimize the surgeon's mental workload during adjustment, a smart screwdriver (SSD) has been developed being able to automatically adjust the screws.

The basic idea of the smart screwdriver is to have a system providing an automatic transfer of the planned data to the cutting guide similar to a robotic system, but with the actuators separated from the kinematic. The use of the SSD is as simple as follows: After planning of the intervention and rigid fixation of the cutting guide on the bone, the surgeon simply connects sequentially the screwdriver to all screws of the cutting guide.

To further maximise the ease of use and to avoid a mix-up of different screws, an identification means has been integrated into the positioning screws as well as into the smart screwdriver. For an automated identification of the screws different technologies have been analysed as position tracking, optical recognition or wired/wireless electronics.

A first prototype without screw identification has been used successfully on 4 cadaver knees. All guide positions could be adjusted automatically using the SSD. However, the absence of screw identification required that the surgeon follows indications given by the computer to turn screws sequentially.

A second prototype of the smart screwdriver has successfully been built up and is able to identify the different positioning screws in less than 1s with high reliability. The identification is realised as inductive coupling of different small resonance circuits that are integrated into the screw heads and the screwdrivers tip.

To adjust the cutting guide from neutral to the planned position, the screws have to be adjusted by 5 mm in average. The rotational speed of the current SSD implementation is 2 rounds per second, resulting in a mean time of about 3.5 s for each screw adjustment. The rotational accuracy of the screwdriver is ±5°. Taking into account a thread of the positioning screws of 0.7 mm, the theoretical translational error is about ±0.01 mm. Looking at the angular accuracy, the maximum distance of the screws of the current setup of the cutting block of 15 mm results in an angular error of less than ±0.05°.