MRes Projects Completed

“Accurate dense 3D reconstruction in minimally invasive surgery” and “Designing a framework for planning and assessment of knee replacement surgery”.

The first year of Hamlyn Centre’s new MRes course in Medical Robotics and Image Guided Intervention has come to an end.  The first group of MRes students had their viva exams in the past week. 

One project involved the accurate dense 3D reconstruction in minimally invasive surgery. Minimally invasive surgery has transformed the conventional practice of surgical operations over the past two decades. However, accompanied difficulties such as indirect 2D visualisation and lack of depth information and haptic feedback highly limit its merits. To overcome these problems, this project looked into registering the pre-operative data and intra-operative data to provide surgical guidance for surgeons. This registration relies on accurate dense 3D reconstruction and devising an optimisation scheme for tissue structure and camera motion recovery.

Another project seeked to improve the current framework for computer assisted knee replacement surgery by the novel introduction of a handheld Metris MCA II 7-axis laser scanner to collect intra-operative data from the patient knee and developing a method for the registration of these data to the pre-operative CT scan. An articulated statistical shape model was made using training shapes taken from a knee phantom and, following registration with the laser scan, this was used to instantiate a digital reconstruction of the knee. The investigation considered a framework for optimal scan planning whereby the anatomical regions of the knee are categorised according to their statistical importance in order to instantiate the most accurate patient-specific model. Finally, the deformation of the knee was modelled after a realistic simulation of both a total knee replacement procedure and a chondroplasty procedure. The results demonstrated the ability of the laser scanner to accurately reconstruct the cutting contours along the knee, thus providing a useful analytical tool for the assessment of the surgical procedure whilst increasing the ease and efficacy of data collection and minimising patient trauma