PRESENTATION OF DIGITIZED AND POSTPROCESSED PORTAL FILMS ON A COMPUTER MONITOR COMPARED TO CONVENTIONAL FILMS PRESENTED ON A LIGHT-BOX

Purpose: Judgement of image quality and detail recognition of digitize d and post-processed portal films presented on a computer monitor comp ared to the present standard, conventional portal films presented on a light-box. Material and Methods: Conventional portal films of 3 diffe rent tumor sites (10 pelvis, 10 cranium, 10 vertebral column) were pre sented to a panel of 8 observers in 3 different manners: conventional film presented on a light-box (Conv), digitized non-post-processed ima ges (Dig-l) and digitized post-processed images (Dig-2) presented on a high-resolution computer monitor. Subjective judgement of image quali ty, detail recognition anti time requirement of conventional films com pared to monitor presentation were evaluated using a 5-scaled question naire (from 1 = much better to 5 = much worse). Furthermore the observ ers had to point out predefined anatomical bony structures on the conv entional films (Conv) as well as on the digitized post-processed image s (Dig-2). Standard deviations of the landmarks outlined by 10 differe nt observers were used as a criterion of objective detail recognition (Figure 1). Results: Image quality of digitized post-processed images presented on the computer monitor was judged statistical significant b etter than of conventional films (pelvis 78%, vertebral column 62%, cr anium 45% better) (Figure 3). Similar results were found for compariso n of detail recognition: digitized post-processed images were scored b etter for pelvis in 81%, for vertebral column in 57%, for cranium in 4 0% (Figure 4, Table 1). Most benefit from portal film enhancement was found for pelvic images, where portal films are known to be of poor im age duality (Figure 2). In contrast image quality of non-processed dig ital images compared to conventional films was graded worse (pelvis 69 %, vertebral column 53%, cranium 71% worse) (Figure 4). Digital post-p rocessed images were especially for the pelvis judged to require less time (pelvis 68%, vertebral column 26%, cranium 8% less time requireme nt) (Figure 5). For the pelvis a statistical significant decrease of s tandard deviations was found for Dig-2 compared to conventional films, indicating an objective increase of image quality and detail recognit ion (Table 2). In case of vertebral column and cranium no significant differences were evaluated (Table 3). Conclusions: Digitized enhanced portal films presented on a computer monitor resulted in a quicker ass essment and equal to better image quality as well as detail recognitio n compared to conventional films. Non-processed digitized images were judged to be of less image quality.