E A Krupinski1, H Roehrig. 1. Department of Radiology, University of Arizona, Tucson 85724, USA.
Abstract
RATIONALE AND OBJECTIVES: The purpose of this study was to determine whether perceptual linearization of the tone scale affects the detection and visual search behaviors of radiologists searching mammograms for masses and microcalcifications. A perceptually linearized display is designed to match the capabilities of the human visual system more closely than a nonlinearized display. MATERIALS AND METHODS: Six radiologists viewed 50 pairs of mammograms, once on a perceptually linearized cathode-ray tube (CRT) monitor and once on a non-linearized CRT monitor. Eye position also was recorded as the observers searched the images for masses and microcalcifications. RESULTS: Observer performance was significantly (P = .003) better with the perceptually linearized display. Dwell times associated with true-negative decisions were significantly longer with use of the nonlinearized display. The number of fixation clusters generated during search was also greater with use of the nonlinearized display for the lesion-free images. CONCLUSION: A perceptually linearized display yields better detection performance and a more efficient visual search. Perceptually linearized displays should be used for reading radiographs displayed on CRT monitors.
RATIONALE AND OBJECTIVES: The purpose of this study was to determine whether perceptual linearization of the tone scale affects the detection and visual search behaviors of radiologists searching mammograms for masses and microcalcifications. A perceptually linearized display is designed to match the capabilities of the human visual system more closely than a nonlinearized display. MATERIALS AND METHODS: Six radiologists viewed 50 pairs of mammograms, once on a perceptually linearized cathode-ray tube (CRT) monitor and once on a non-linearized CRT monitor. Eye position also was recorded as the observers searched the images for masses and microcalcifications. RESULTS: Observer performance was significantly (P = .003) better with the perceptually linearized display. Dwell times associated with true-negative decisions were significantly longer with use of the nonlinearized display. The number of fixation clusters generated during search was also greater with use of the nonlinearized display for the lesion-free images. CONCLUSION: A perceptually linearized display yields better detection performance and a more efficient visual search. Perceptually linearized displays should be used for reading radiographs displayed on CRT monitors.
Authors: Elizabeth A Krupinski; Louis D Silverstein; Syed F Hashmi; Anna R Graham; Ronald S Weinstein; Hans Roehrig Journal: J Digit Imaging Date: 2012-12 Impact factor: 4.056
Authors: Xiao Hui Wang; Janet E Durick; Amy Lu; David L Herbert; Saraswathi K Golla; Kristin Foley; C Samia Piracha; Dilip D Shinde; Betty E Shindel; Carl R Fuhrman; Cynthia A Britton; Diane C Strollo; Sherry S Shang; Joan M Lacomis; Walter F Good Journal: J Digit Imaging Date: 2007-09-15 Impact factor: 4.056
Authors: Tarek N Hanna; Matthew E Zygmont; Ryan Peterson; David Theriot; Haris Shekhani; Jamlik-Omari Johnson; Elizabeth A Krupinski Journal: J Am Coll Radiol Date: 2018-02-01 Impact factor: 5.532
Authors: Steve Si Jia Feng; Carl J D'Orsi; Mary S Newell; Rebecca L Seidel; Bhavika Patel; Ioannis Sechopoulos Journal: Med Phys Date: 2014-03 Impact factor: 4.071