作者: David Gur , David A. Rubin , Barry H. Kart , Arleen M. Peterson , Carl R. Fuhrman
DOI: 10.1007/BF03168596
关键词: Two-alternative forced choice 、 Ordinal Scale 、 Image resolution 、 Artificial intelligence 、 Radiographic Image Enhancement 、 Observer (quantum physics) 、 Computer vision 、 Image quality 、 Scale (ratio) 、 Rating scale 、 Computer science 、 Radiological and Ultrasound Technology 、 Radiology Nuclear Medicine and imaging 、 Computer Science Applications
摘要: This study compared a five-category ordinal scale and two-alternative forced-choice subjective rating of image quality preferences in multiabnormality environment. 140 pairs laser-printed posteroanterior (PA) chest images were evaluated twice by three radiologists who asked to select during side-by-side review which each pair was the “better” one for determination presence or absence specific abnormalities. Each included (the digitized film at 100 μm pixel resolution laser printed onto film) highly compressed (∼60∶1) decompressed version that film. Ratings performed once with scale. The selection process significantly affected used. “comparable” “equivalent diagnosis” category used 88.5% ratings When using approach, noncompressed selected 66.8% time as being images. resulted lower ability detect small differences perceived between when is Observer behavior can be type question Observers are sensitive presentation review.