To make ‘keyhole’ heart operations safer and easier to perform
Minimally invasive heart surgery, in which the surgeon operates via tiny endoscopic cameras and specialised instruments, cuts the mortality risk and leads to faster, less painful recovery than conventional open chest procedures.
But the technique limits the surgeon’s view of the patient, said Dr Daniel Rueckert, senior lecturer in computing at
Traditional image-guided robotic surgery is more often used for operations on the bones and joints or the brain, which do not move during the procedure.
But when operating on the heart or the blood vessels attached to it, such as during bypass surgery, the patient’s heart is continually beating. Tracking a moving target makes it even more difficult for surgeons operating with an already restricted view.
Rueckert and his team are working with researchers at St Mary’s Hospital and the Centre for Medical Image Computing at UCL to develop a system to match the 2D endoscopic images seen by the surgeon, with a detailed model of the moving, 3D body tissue.
‘One of the goals of the project is to provide guidance to the surgeon by pointing out certain anatomical features in the scene which they are either trying to navigate around or attempting to avoid; if these features were damaged during surgery it might be dangerous for the patient,’ said Rueckert.
3D images taken before the operation in routine MRI or CT scans will be superimposed on to the live 2D video images, to allow the surgeon to more accurately locate and track target tissue structures.
Techniques developed by the team in previous research projects will be used to detect particular features of interest within the endoscopic image — and these will be used to align the pre-operative model with the live video.
Although the main aim of the EPSRC-funded project, due to begin this summer, is for heart surgery, the technology could also be used to improve other procedures on soft tissue, such as liver cancer treatment.
The system will have built-in checks to ensure it is completely accurate. If it finds certain anatomical features are not where it expects them to be, it will alert the surgeon that it is currently unable to provide guidance, said Rueckert.
The researchers plan to build a phantom heart on which to extensively test the image guide using the robotic surgery system at St Mary’s Hospital. The tests are expected to take place in the first half of 2006.
By the summer of 2007 they hope to have produced a demonstrator system.