Image Quality Assessment and EquipmentOptimisation with Automated Phantom Evaluation in Full Field Digital Mammography (FFDM)

摘要

The work investigates the feasibility of using a commercial software performing automatic threshold-detail-image analysis for absolute image quality evaluation (phantom CDMAM ver. 4.0 + CDMAM Analyser ver. 1.1). The software proposes an overall image quality index IQF_(inv). The quantities IQF_(inv), contrast-to-noise-ratio (CNR), threshold contrast and average glandular dose (AGD) are studied for a wide range of exposure settings in three full field digital mammographic units of different manufacturers. IQF_(inv) correlates linearly with CNR and exhibits a better correlation to threshold contrast than CNR itself in all units, turning out to be a reliable parameter for image quality assessment. At comparable AGD and exposure settings, all systems exhibit equivalent image quality. Using the automatic exposure control (AEC) all units achieve, for thin and medium breasts, a higher image quality than requested by the European Guidelines, whereas only in one unit the image quality remains above the desired level for thick breasts.