Four-dimensional computed tomography (4DCT: three dimensional + time) allows to measure individual bone position over a period of time usually during motion. This method has been found useful in studying the joints around the wrist as dynamic instabilities are difficult to detect during static CT scans while they can be diagnosed using a 4DCT scan [1]–[3]. For the foot, the PedCAT system (Curvebeam, Warrington, USA) has been developed to study the foot bones under full weight bearing, however its use is limited to static images. On the contrary, dynamic measurements of the foot kinematics using skin markers can only describe motion of foot segments and not of individual bones. However, the ability to measure individual bone kinematics during gait is of paramount importance as such detailed information could be used to detect instabilities, to evaluate the effect of joint degeneration, to help in pre-operative planning as well as in post-operative evaluation.

The overall gait kinematics of two healthy volunteers were measured in a gait analysis lab (Movement Analysis Lab Leuven, Belgium) using a detailed foot-model (Oxford foot model, [4]). The measured plantar-dorsiflexion and in-eversion were used to manipulate their foot during a 4D CT acquisition. The manipulation was performed through a custom made foot manipulator that controls the position and orientation of the foot bed according to input kinematics. The manipulator was compatible with the 4D CT Scanner (Aquilion One, Toshiba, JP), and a sequence of CT scans (37 CT scans over 10 seconds with 320 slices for each scan and a slice thickness of 0.5 mm) was generated over the duration of the simulation. The position of the individual bones was determined using an automatic segmentation routine after which the kinematics of individual foot bones were calculated. To do so, three landmarks were tracked on each bone over time allowing to construct bone-specific coordinate frames. The motion of the foot bed was compared against the calculated kinematics of the tibia-calcaneus as the angles between these two bones are captured with skin markers.

There is high repeatability between the imposed plantar/dorsiflexion and inversion/eversion and the calculated. Although the internal/external rotation was not imposed, the calculated kinematics follow the same pattern as the measured in the gait-analysis lab. Based on the validation of the tibia-calcaneus, the kinematics were also calculated between four other joints: tibia-talar, talar-calcaneus, calcaneus-cuboid and talar-navicular. Repeatable measurements of individual foot bone motion were obtained for both volunteers.

The use of 4D CT-scanning in combination with a foot manipulator can provide more detailed information than skin marker-based gait-analysis e.g. for the study of the the tibia-talar joint. In the future, the foot manipulator will be tested for its sensitivity for specific pathologies (e.g. metatarsal coalition) and will be further developed to better resemble a real-life stance phase of gait (i.e. to include isolated heel contact and toe off).