Preprocessing for improving CAD scheme performance for microcalcifications detection based on mammography imaging quality parameters Homero Schiabel* a , Marcelo A. C. Vieira a , Liliane Ventura a a Dept. Electrical Engineering, EESC – Universidade de S. Paulo, Av. Trabalhador Saocarlense 400, Sao Carlos (SP), Brasil, 13566-590 ABSTRACT Database characteristics can affect significantly the performance of a mammography CAD scheme. Hence adequate performance comparison among different CAD schemes is not suitable since a single scheme could present different results depending on the set of chosen cases. Images in database should follow a set of quality criteria, since the imaging process up to digital file. CAD schemes can not be developed without a database used to test their efficacy, but each database with particular characteristics may influence on the processing scheme performance. A possible solution could be using information on the imaging equipment characteristics. This work describes a preprocessing in order to “compensate” the image degradation during the acquisition steps, assuring a better “uniformity” relative to the images quality. Thus, poor quality images would be restored, providing therefore some independence on the images source to CAD schemes and allowing to reach the better possible performance. Tests performed with mammography images sets reported a 14% increase in sensitivity for microcalcifications detection. Although this result was followed by a little increase in false positive rates, simple changes in techniques parameters can provide the same improvement but with a reduction of the false positive detections. Keywords: Computer-aided diagnosis in Mammography, microcalcifications detection, mammography images database 1. INTRODUCTION From the development of more sophisticated radiographic systems along the 1990’s, an increased interest has been registered regarding automatic schemes for aiding the diagnosis in radiology. Among the main developed techniques driven to mammography, the main attention was given to those intended to detect and/or classify microcalcifications [1-5] , suspicious masses and tumors [2,6,7] , besides preprocessing techniques designed to enhance mammographic images contrast [8-11] . The basic procedures associated to the mammographic image computerized processing generally begin with the original mammograms digitization and the subsequent application of processing techniques on the digital image. Only in direct digital mammography equipment the mammogram digitization is obviously not required since the digital image is directly acquired by the system detectors plate [12,13] . Computer-Aided Detection (CAD) schemes have the purpose of providing a “second opinion” to the radiologist, aiding in the detection of suspicious lesions from a mammogram, and investigating them from their benign or malignant characteristics. Therefore, such schemes main purpose is the improvement of the mammographic examination efficacy, with simultaneous reduction in the amount of diagnosis errors – reflected by overlooking or unneeded biopsies [1,14-16] . In 1998, FDA has approved the first commercial CAD scheme for clinical use in mammography – the ImageChecker from R2 Technology, Inc. (Los Altos, EUA). Estimations have pointed out more than 1,500 CAD schemes being used currently in hospitals and radiological clinics in the USA as aiding in breast cancer screening [17] . Recent works have shown a significant increase in radiologists performance when assisted by a CAD scheme. Freer & Ulissey [18] have evaluated their own diagnoses in mammography when using a CAD scheme for one year in the clinical routine. During this period, they have investigated 12,860 mammograms, providing firstly the diagnosis without the CAD aiding and, then, reviewing the same diagnosis based on the result provided by the automatic scheme. The results have shown an increase of 19.5% in the number of breast cancer cases correctly detected when assisted by the CAD scheme, without a significant increase in the number of unnecessary biopsies. *homero@sel.eesc.usp.br; phone #55 16 3373-9365; fax #55 16 3373-9372 Medical Imaging 2009: Computer-Aided Diagnosis, edited by Nico Karssemeijer, Maryellen L. Giger Proc. of SPIE Vol. 7260, 72602G · © 2009 SPIE · CCC code: 1605-7422/09/$18 · doi: 10.1117/12.812015 Proc. of SPIE Vol. 7260 72602G-1