View the latest medical engineering news

[January 2007] Bioengineering Sciences Research Group, School of Engineering Sciences, University of Southampton

Increasing activity in our society has led to a growing number of joint injuries and disorders, which can eventually necessitate joint replacement surgery. Today, there are approximately 540,000 knee replacement operations per year in the EU alone and these numbers are only expected to increase further. Unfortunately, one in twenty patients over 60 and one in ten patients under 60 will require a revision procedure within ten years of their total knee replacement (TKR). Subsequent surgeries are problematic due to depleted bone stock in which to implant the components. This is of increasing concern as the patient demographic is becoming focused on younger (under 60) and more active patients, who will also place higher kinematic demands on the device post-operatively. This means that the most important features of a TKR are post-operative function and implant longevity.

The major challenges of TKR surgery are the selection of an appropriate prosthesis (both design and size), and optimal positioning of the device based on the patient’s individual characteristics. At present, surgeons base their decisions on personal experience and implant geometry. Variability in patient outcome is highly dependent upon the experience and skill of the individual surgeon, and there are at present no knowledge-based systems available to assist during the planning of the operation that takes patient-specific data into account. Image Guided Surgery (IGS) systems are becoming increasingly widespread and allow some anatomical data to be collected from the patient either pre- or intra-operatively in order to prepare a surgical plan. However, the implications are that the surgical plan is still based on the generic, geometry-based technique recommended by the manufacturer and on the surgeon’s experience. In the worst case, the prosthesis could be accurately positioned, but in a less than optimal position. In addition to this, the current practices give no indication of the post-operative function of the replaced joint, although it is vital to the surgical outcome.

In order to improve upon the current failure rate of TKR and to improve their longevity and functional performance, it is necessary to enhance the prosthesis selection/sizing and positioning process used either pre-or intra-operatively by surgeons, in order to achieve the optimal configuration for the patient. With this in mind the ultimate goals of the DeSSOS project are the following: 

• The ability to extract anatomical landmarks (such as ankle centre, knee flexion axis or ligament ankle points) from disparate forms of patient data (CT, MRI, IGS, etc). In particular identify and extract geometric information related to the main ligament structures. 
• The ability to predict the kinematic function of a replacement knee based on a given surgical plan for a patient, in a timescale appropriate for use either pre- or intra-operatively. 
• The ability to determine the most suitable size and orientation of the prosthetic components in order to obtain the optimal performance for a particular patient, in a timescale appropriate for use either pre- or intra-operatively.

Through our developments, we aim to reduce the number of revision operations and improve overall post-operative function for the patient.

To address this objective DeSSOS consists of nine partners; five research institutions: University of Southampton (lead partner), Charité (Center for Musculoskeletal Research), University of Leiden, University of Zaragoza, and Zuse Institute Berlin (ZIB); three industrial partners: DePuy International, Finsbury Instruments, and ESI Group; and one management partner: Pera Innovation. The flow of work is shown in figure 1.

	Figure 1. DeSSOS work flowchart

Anatomical landmarks, such as femoral head centre, ankle centre and knee flexion axis, are gathered either pre-operatively using CT or MRI information, or intra-operatively using an IGS system. A generic model of a lower limb is then morphed into a patient specific musculoskeletal model using this data. A force prediction module adapts patient joint contact and muscle forces to be used as inputs for the patient specific lower limb FE model with an implanted TKR (figure 2). The FE model outputs post-operative performance prediction for the given prosthesis placement. This will allow the surgeon to decide if the TKR is optimally aligned and plan the surgery accordingly. These models will include realistic ligaments with complex nonlinear material behaviour.

In order to calculate the forces, and hence the kinematics, a number of assumptions must be made. Each assumption, along with any measured input device, will have a degree of uncertainty associated with it. Probabilistic approaches will account for the uncertainty and predict a range of kinematics a patient would be likely to experience for a given activity. The process will be assessed, tested and validated as far as possible against the individual kinematics of a number of patients.

	Figure 2. Finite element model of a generic lower limb with implanted TKR

The DeSSOS project will provide a new knowledge-based system to assist during the planning and performance of the operation that takes into account patient-specific data, including anatomical landmarks and ligament positions. This system will also be implemented to carry out preclinical testing of new devices, as well as for surgeon training purposes. This will allow new surgeons to virtually plan and carry out surgery and see the effects of small changes in prosthesis placement on the post-operative performance.

This work is funded by the EU (IST-2004-27252)

For further information contact Prof. Mark Taylor (M.Taylor@soton.ac.uk)

 

Back to top