Aims

The use of technology to assess balance and alignment during total knee surgery can provide an overload of numerical data to the surgeon. Meanwhile, this quantification holds the potential to clarify and guide the surgeon through the surgical decision process when selecting the appropriate bone recut or soft tissue adjustment when balancing a total knee. Therefore, this paper evaluates the potential of deploying supervised machine learning (ML) models to select a surgical correction based on patient-specific intra-operative assessments.

The February 2015 Knee Roundup. 360 . looks at: Intra-operative sensors for knee balance; Mobile bearing no advantage; Death and knee replacement: a falling phenomenon; The swings and roundabouts of unicompartmental arthroplasty; Regulation, implants and innovation; The weight of arthroplasty responsibility!; BMI in arthroplasty

The aim of this prospective multicentre study was to report the patient satisfaction after total knee replacement (TKR), undertaken with the aid of intra-operative sensors, and to compare these results with previous studies. A total of 135 patients undergoing TKR were included in the study. The soft-tissue balance of each TKR was quantified intra-operatively by the sensor, and 18 (13%) were found to be unbalanced. A total of 113 patients (96.7%) in the balanced group and 15 (82.1%) in the unbalanced group were satisfied or very satisfied one year post-operatively (p = 0.043). . A review of the literature identified no previous study with a mean level of satisfaction that was greater than the reported level of satisfaction of the balanced TKR group in this study. Ensuring soft-tissue balance by using intra-operative sensors during TKR may improve satisfaction. . Cite this article: Bone Joint J 2014;96-B:1333–8

Over the last decade Computer Assisted Orthopaedic Surgery (CAOS) has emerged particularly in the area of minimally invasive Uni-compartmental Knee Replacement (UKR) surgery. Image registration is an important aspect in all computer assisted surgery including Neurosurgery, Cranio-maxillofacial surgery and Orthopaedics. It is possible for example to visualise the patient's medial or lateral condyle on the tibia in the pre-operated CT scan as well as to locate the same points on the actual patient during surgery using intra-operative sensors or probes. However their spatial correspondence remains unknown until image registration is achieved. Image registration process generates this relationship and allows the surgeon to visualise the 3D pre-operative scan data in-relation to the patient's anatomy in the operating theatre. Current image registration for most CAOS applications is achieved through probing along the articulating surface of the femur and tibial plateau and using these digitised points to form a rigid body which is then fitted to the pre-operative scan data using a best fit type minimisation. However, the probe approach is time consuming which often takes 10–15 minutes to complete and therefore costly. Thus the rationale for this study was to develop a new, cost effective, contactless, automated registration method which would entail much lesser time to produce the rigid body model in theatre from the ends of the exposed bones. This can be achieved by taking 3D scans intra-operatively using a Laser Displacement Sensor. A number of techniques using hand held and automated 3D Laser scanners for acquiring geometry of non-reflective objects have been developed and used to scan the surface geometry of a porcine femur with four holes drilled in it. The distances between the holes and the geometry of the bone were measured using digital vernier callipers as well as measurements acquired from the 3D scans. These distances were measured in an open source package MESHLAB version 1.3.2 used for the interpretation, post-processing and analysis of the 3D meshes. Absolute errors ranging from of 0.1 mm to 0.4 mm and the absolute percentage errors ranging from 0.48% to 0.75% were found. Additionally, a pre-calibrated dental model was scanned using a 650 nm FARO™ Laser arm using the global surface registration approach in Geomagic Qualify package and our 3D Laser scanner. Results indicate an average measurement error of 0.16 mm, with deviations ranging from 0.12mm to −0.13 mm and a standard deviation of 0.2 mm. We demonstrated that by acquiring multiple scans of the targets, complete 3D models along with their surface texture can be developed. The overall scanning process, including time required for the post-processing of the data requires less than 20 minutes and is a cost-efficient approach. Moreover, the majority of that time was used in post processing the acquired data which could be potentially reduced through the use of bespoke application software. This project has provided proof of concept for a new automated, non-invasive and cost efficient registration technique with the potential of providing a quantitative assessment of the articular cartilage integrity during lower limb arthroplasty