Introduction the link between low bone mineral density (BMD) leading to greater fracture risk is well established in the literature; what is not fully understood is the impact of total knee revisions (rTKR) and cone implantation on BMD. This is important due to the increasing fracture risk associated with reductions in BMD. This feasibility study investigated a new type of Stryker cone for rTKR patients, and its impact on BMD utilising different imaging technologies and providing recommendations to be implemented for a full follow up trial.
Method a systematic review was conducted to investigate total knee replacement (TKR) and rTKR on BMD results to establish known reported BMD changes after surgery, and to highlight the knee regions investigated. A bovine study was then conducted in order to test the different setup imaging technologies and possible analysis of the cones. Additionally, a novel piece of 3D SHAPER hip software was utilised to investigate bone changes in the hip across three groups (TKR, rTKR, and controls) which could then be compared to the main BMD changes or used as an alternative to the other imaging options. The main study involved recruiting 37 participants all undergoing rTKR to either a cone or non cone group, with all participants undergoing a series of scans via: CT scans (only at six months), DXA and x ray at intervals of pre op, six weeks, three, six and 12 months. Additionally, all participants completed questionnaires on mental health, lower extremity functionality, and quality of life. In addition to BMD investigation, hip and knee alignment was also explored at pre and post op intervals, as well as pixel density changes, both utilising long leg x ray imaging.
Results Systematic review results reported 2,431 papers, of which 27 studies were included, across all the studies BMD losses appeared greatest at 12 months. The bovine study helped develop the imaging and analysis required for the main study. The 3D SHAPER ability to be applied to hip DXA imaging showed promise; which was reflected in the control, rTKR and TKR data. The development of different imaging technologies have potential in moving forward into a full trial. Recommendations would include: utilising DXA imaging as the main modality, given its gold standard for BMD changes and its consistency when using a standardised positioning protocol and ROI placement. Long leg x-
3
ray imaging to be used to investigate alignment and pixel density changes, as this imaging is convenient as part of routine follow-up care, although the inclusion of a step wedge within all long leg images would be required to allow pixel density standardisation for investigating in-growth. Finally, the CT imaging could not determine ingrowth in this feasibility study, and therefore should not be utilised in the full study. For the main feasibility study results, 35 participants attended pre--op, 26 attended six weeks and three months, at six months 25 attended, and 22 at 12 months. Results show rTKR is associated with lower BMD in the tibial and femoral stems, and in the medial tibial condyle, and associated with increases beyond the tibial and femoral stems, in both groups. The main difference is in lateral tibial condyle where there are associated increases in BMD in the cone group, and losses reported in the non--cone group. The questionnaire results show a favourable impact for rTKR, with reductions in depression, anxiety, and increases in functionality post--surgery, with the cone group reporting greater changes, although not statistically significant between groups. Alignment analysis shows little difference between.

Objectives: to investigated the accuracy of computer-aided detection (CAD) software in musculoskeletal projection radiography via a systematic review. Key findings: Following selection screening, eligible studies were assessed for bias, and had their study characteristics extracted resulting in 22 studies being included. of these 22 three studies had tested their CAD software in a clinical setting; the first study investigated vertebral fractures, reporting a sensitivity score of 69.3% with CAD, compared to 59.8% sensitivity without CAD. The second study tested dental caries diagnosis producing a sensitivity score of 68.8% and specificity of 94.1% with CAD, compared to sensitivity of 39.3% and specificity of 96.7% without CAD. The third indicated osteoporotic cases based on CAD, resulting in 100% sensitivity and 81.3% specificity. Conclusion: the current evidence reported shows a lack of development into the clinical testing phase; however the research does show future promise in the variation of different CAD systems.

The link between low bone mineral density (BMD) scores leading to greater fracture risk is well established in the literature; what is not fully understood is the impact of total knee replacements/revisions or arthroplasties on BMD levels. This literature review attempts to answer this question. Several different databases using specific key terms were searched, with additional papers retrieved via bibliographic review. Based on the available evidence, total knee replacements/revisions and arthroplasties lower BMD and thus increase fracture risk. This review also addresses the possible implications of this research and possible options to reduce this risk.

Corners hold vital information about size, shape and morphology of a vertebra in an x-ray image, and recent literature (Al-Arif et al. 2015) [1,2] has shown promising performance for detecting vertebral corners using a Hough forest-based architecture. To provide spatial context, this method generates a set of 12 patches around a vertebra and uses a machine learning approach to predict corners of a vertebral body through a voting process. In this paper, we extend this framework in terms of patch generation and prediction methods. During patch generation, the square region of interest has been replaced with data-driven rectangular and trapezoidal region of interest which better aligns the patches to the vertebral body geometry, resulting in more discriminative feature vectors. The corner estimation or the prediction stage has been improved by utilising more efficient voting process using a single kernel density estimation. In addition, advanced and more complex feature vectors are introduced. We also present a thorough evaluation of the framework with different patch generation methods, forest training mechanisms and prediction methods. In order to compare the performance of this framework with a more general method, a novel multi-scale Harris corner detector-based approach is introduced that incorporates a spatial prior through a naive Bayes method. All these methods have been tested on a dataset of 90 X-ray images and achieved an average corner localization error of 2.01 mm, representing a 33% improvement in localization accuracy compared to the previous state-of-the-art method (Al-Arif et al. 2015) [2].1

Developing CSPINE CAD through machine learning algorithms: Inter-operator precision errors of user inputs. Liverpool 29 June – 1 July 2015. UKRC conference proceedings. P63

Student radiographer perceptions of using CSPINE CAD software to assist cervical spine image interpretation and diagnosis. Liverpool 29 June – 1 July 2015. UKRC conference proceedings. P108: P008