Reply: A plea for confidence intervals and consideration of generalizability in diagnostic studies

摘要

Sir, We read with interest the recent article by Klo ¨ ppel andncolleagues (2008) on automatic classification of MR scans innAlzheimer’s disease and the coverage on the BBC’s website.nIn particular, we were concerned that the claim on the websitenthat ‘computers ‘‘spot Alzheimer’s fast’’ ’ cannot be justified onnthe basis of this research. Although the authors do not makenthis claim in their paper, in our opinion they do not pay enoughnattention to two limitations of their work, which may havenencouraged the media over-interpretation of their findings.