Objective: To determine how Larsen scores from digitized X-rays compare to those from film originals.
Methods: A hundred sets of radiographs of patients recruited with early rheumatoid arthritis (RA) were assessed using the Larsen scoring system. Digitized copies of these sets were then viewed on a computer screen and scored according to Larsen in a random order. The quality of the digitized image was also recorded. For each set of X-rays, the signed difference between the score from film and the score from the digitized images was calculated.
Results: A total of 95% of the digitized X-ray sets were scored successfully; 5% were not scored due to the images being unreadable. The mean difference between the two sets of scores was -1.2 (95% CI [-2.06, -0.37]). There was no trend in the difference with respect to the mean of the two scores (P>0.1).
Conclusion: The Larsen scoring of digitized X-ray images has been validated.