Computerized evaluation of mammographic image quality using phantom images Geoffrey Dougherty* Department of Radiologic Sciences, Faculty of Allied Health Sciences, Kuwait University, PO Box 31470, 90805 Sulaibikhat, Kuwait Received 7 January 1998; accepted 5 August 1998 Abstract A simple, quick and computerized method for quantitatively evaluating the image quality of mammography phantom images has been developed. Images of the American College of Radiology (ACR) mammographic accreditation phantoms were acquired under different X- ray techniques, scored and ranked subjectively by ®ve expert readers, and digitized for quantitative analysis. The contrast and signal-to-noise (contrast-to-noise) ratios of the main nodule and microcalci®cation group were obtained accurately and reproducibly using an image processing protocol. The contrast values were successful at discriminating differences in image quality due to variations in scatter conditions (as a result of different kVp's, and the presence or absence of an acrylic scatterer and/or a moving Bucky grid). They were more precise, reproducible and sensitive than the ACR score. In particular, the contrast of the main nodule was highly correlated (r s  0.988: p , 0.001) with the ranking of image quality by our panel of expert readers. q 1998 Elsevier Science Ltd. All rights reserved Keywords: Mammography; Mammographic phantom; Image quality; Conspicuity 1. Introduction Worldwide, breast cancer is one of the most serious health threats facing women. In the United States cancer of the breast is the most frequently diagnosed malignancy [1] and is a leading cause of cancer mortality among women [2]. At present, mammography remains the sole imaging modality with proven capability of detecting small, clinically occult breast lesions in asymptomatic women [3]. The medical value of mammographic screening for breast cancer has been demonstrated in a number of trials [4±11], albeit with some reservations on age and frequency of screening [12,13]. Survival and recovery depend on early diagnosis, and hence the image quality of the mammo- graphic system is of crucial importance. In particular, the portrayal of ®ne detail and the improvement of contrast with surrounding ®broglandular tissue are the keys to the visua- lization of early or subtle abnormalities [14,15]. Although mammography is more accurate than any other currently available modality, limitations in sensitivity (typically, 80±90%) and speci®city (typically 15±30%) are a concern. Meticulous attention to detail in the mammographic assessment is required since even a very slight decrease in image quality could result in the loss of diagnostically use- ful information. High quality mammographic images depend on all the factors within the imaging chainÐincluding the X-ray equipment, the screen±®lm combination and its degree of contact [16], and the processing. Technical image quality is conventionally described in terms of the contrast, sharpness and noise properties of the image formation process [17], since these elements can be determined on a more or less independent basis. However, the effects of these three image properties on the visibility of ®ne mammographic detail are interrelated, and their relative importance in the depiction of subtle but diagnostically signi®cant abnormalities depends on the target of interest. Consequently breast phantoms, which contain structures that mimic clinically relevant fea- tures such as soft-tissue masses and calci®cations, are rou- tinely used to monitor the performance of a mammographic imaging system and assess global image quality. Evaluation using a phantom is convenient to conduct in a standardized manner since patient radiation dose is not a factor, although there are concerns that these tests may not adequately simu- late the radiologists' search for speci®c anatomic details against highly complex backgrounds [18,19]. A number of different phantoms are commercially avail- able [20,21]. They contain a range of structures some of which are more sensitive to resolution related factors 0895-6111/98/$ - see front matter q 1998 Elsevier Science Ltd. All rights reserved PII: S0895-6111(98)00043-3 * Corresponding author. Fax: +965-4833662; e-mail: geoff@hsc.kuniv. edu.kw PERGAMON Computerized Medical Imaging and Graphics 22 (1998) 365±373 Computerized Medical Imaging and Graphics