Introduction: Digital X-rays have become increasingly prevalent in Hospitals throughout the UK and Ireland in the past 10 years. We have devised a semi quantitative analysis of digital radiographs that measures the extent of healing across the fracture gap.

Methods: 48 CD 1 mice underwent a femoral fracture and subsequent fixation with an external fixator. A standardised radiograph was taken. A radiographic analysis was carried out. For each radiograph taken a pixel density graph was generated at five individual points across the fracture gap, along the longitudinal axis of the femur.

A stastical analysis of intra and inter-observer variability was tested using the linearly-weighted kappa statistic for each of the 240 pixel density graphs taken and for the summation total in the 48 radiographs.

Results: For the individual pixel density graphs we expected an agreement of 67.82%. An agreement of 95.42% was recorded showing a kappa statistic of 0.8576 and a standard error of 0.0531.

On analysis of the summation scores we expected an agreement of 75.54% and observed an actual agreement of 96.30%. This showed a kappa statistic of 0.8545 and a standard error of 0.0849.

Conclusion: The results are very similar in the two analyses and indicate excellent agreements. As a result we offer a radiographic, semi-quantative analysis of bone healing across a fracture gap that is highly reproducible. Thus it has the potential for application to future research in this field and possibly to clinical practice with the increased use of digital radiographs in hospital departments