The knee ops diagnosed by computer

A COMPUTER program could help cut costly or unnecessary surgery on patients with knee replacements.

Scientists have developed a diagnostic tool to counter the soaring rise of knee replacement operations.

More than 90,000 are performed in the UK each year but this is expected to rise by more than 600 per cent by 2030 due to an ageing population, obesity and younger people having the surgery.

Patients often require a second knee replacement or ‘revision’, which can cost up to four times as much as the original surgery and bring increased risk of infection and failure.

But researchers at the University of Bath have developed a semi-automated computer program to assess the condition of a replacement, a task usually performed ‘by eye’ by surgeons. It gives a score for radiolucency – the region surrounding a hip or knee replacement which is dark on an X-ray and can progressively worsen with loosening – provides a more accurate, independent method to assess the need for extra surgery.

Richie Gill, professor of Healthcare Engineering at the university, said: ‘Until we did the work that we did, all radiolucency lines were looked at by eye by the surgeons. What was interesting to us was the lack of agreement between different surgeons looking at the X-rays.

‘Using a digital computerised tool that accurately identifies patients with progressive pathological radiolucency, showing severely loosening knee replacements, would ensure correct surgi- cal procedures are applied. This will improve outcomes and save money spent on operations which may not ultimately be successful.’

Knee replacements are a highly successful method of relieving the pain and disability of knee osteoarthitis, which usually affects those aged over 40. Studies suggest almost one in two people will suffer from symptomatic osteoarthritis – the degeneration of joint cartilage and underlying bone – of the knee during their lifetime.

The study, published in the Journal of the Royal Society Interface, compared surgeons’ assessments of radiolucency in 38 knee replacement radiographs with those made by the computer program. There was large variation between the surgeons’ results, with total agreement in less than 10 per cent of zones examined, compared with 81.6 per cent in those of the program’s.