Amarpreet Singh Chawla MS, Presenter: Nothing to Disclose

Noriyuki Hashimoto, Abstract Co-Author: Nothing to Disclose

Kazuyuki Kajikawa, Abstract Co-Author: Nothing to Disclose

Thomas Waletzki, Abstract Co-Author: Nothing to Disclose

To compare human observer performance on an LCD at different adaptation levels of the eye and at two different maximum luminance-settings and determine the influence of eye glare.

A capability to render high luminance may not translate into improved diagnostic performance of soft-copy displays in the clinical environment. This is because high luminance can result in significant glare causing eye strain and fatigue. To verify this, human observer performance in detecting objects on a monitor at two different maximum luminance-settings, namely 450cd/m² and 600cd/m², was compared. A 3-megapixel G31-S Eizo monochrome LCD was used for the purpose. The objects consisted of low-contrast periodic and aperiodic patterns. The periodic patterns closely resembled the Standard Target used by the DICOM committee. The objects were embedded in different surroundings to simulate a real clinical situation where there is a local variation in the adaptation level of a radiologist’s eye when viewing a typical medical image. The images were 11-bit, 1536 x 2048 in size. Same set of images was shown at the two calibration settings.

The differences in observer performance between the two calibration settings when viewing objects of different contrasts embedded over the same surrounding level, were negligible. These differences were still less when the objects were brighter than the luminance level of the eye adaptation or when the adaptation level was much higher. Further, it was found that at a higher level of eye adaptation, the observer performance sharply deteriorated when the maximum luminance-setting was increased from 450cd/m² to 600cd/m².

When viewing medical images on a soft-copy display, diagnostic accuracy is influenced by the luminance level to which a display is calibrated. An LCD calibrated to the DICOM GSDF may not be strictly perceptually linearized. A high luminance setting of the monitor may offer little improvement or infact deteriorate diagnostic performance. This can be attributed to higher eye-glare at a higher luminance setting. The glare can be significant even at a typical setup with maximum luminance of 600cd/m².

N.H.,K.K.,T.W.: Noriyuki Hashimoto,Kazuyuki Kajikawa and Thomas Waletzki work for Eizo Nanao Technolgies, USA.

Chawla, A, Hashimoto, N, Kajikawa, K, Waletzki, T, Influence of Display Luminance on Diagnostic Performance: A Psychophysical Study.  Radiological Society of North America 2005 Scientific Assembly and Annual Meeting, November 27 - December 2, 2005 ,Chicago IL. http://archive.rsna.org/2005/4409372.html