Premium
An evaluation of the signal and noise characteristics of four CCD‐based film digitizers
Author(s) -
Hangiandreou N. J.,
O'Connor T. J.,
Felmlee J. P.
Publication year - 1998
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1118/1.598390
Subject(s) - digitization , computer science , laser , noise (video) , detector , teleradiology , image quality , signal (programming language) , optics , optical transfer function , medical imaging , signal to noise ratio (imaging) , artificial intelligence , computer vision , physics , telecommunications , image (mathematics) , health care , telemedicine , economics , programming language , economic growth
Film digitizers are common devices in radiology departments involved with picture archive and communication systems (PACS) and teleradiology. In this paper, we studied the performance of film digitizers based on charge‐coupled device detectors (CCD digitizers), and compared this with the performance of a laser digitizer (the de facto standard). Our focus was on the assessment of signal, noise and useful optical density range performance. A function( L *Δ D ) derived from the Rose model was used to evaluate these parameters in absolute terms, based their predicted ability to detect objects of specific size and optical density difference with respect to background. We studied CCD digitizers from four different vendors and found that none was able to reliably operate up to the maximum density of 3.0 required to digitize plain radiographs, while the laser digitizer was capable of this task. Our analysis also indicated that two of the four CCD digitizers were adequate for digitizing laser‐printed cross‐sectional images in certain cases. Finally, our analysis indicated that digitization of SMPTE pattern films along with visual assessment of the 5% and 95% contrast patches was not sufficient for determining the utility of film digitizers for clinical tasks. Computation of the L *Δ D function provides a useful means of assessing the performance of film digitizers (e.g., for acceptance testing and quality control), and this technique may be adaptable for evaluation of other digital imaging modalities.