TY - JOUR
T1 - Correlative feature analysis on FFDM
AU - Yuan, Yading
AU - Giger, Maryellen L.
AU - Li, Hui
AU - Sennett, Charlene
N1 - Funding Information:
This work was supported in part by US Army Breast Cancer Research Program (BCRP) Predoctoral Traineeship Award (W81XWH-06-1-0726), by United States Public Health Service (USPHS) Grant Nos. CA89452 and P50-CA125183, and by Cancer Center Support Grant (5-P30CA14599). MLG is a stockholder in, and receives royalties from, R2 Technology, Inc (Sunnyvale, CA), a Hologic Company. It is the University of Chicago Conflict of Interest Policy that investigators disclose publicly actually or potential significant financial interest which would reasonably appear to be directly and significantly affected by the research activities.
PY - 2008
Y1 - 2008
N2 - Identifying the corresponding images of a lesion in different views is an essential step in improving the diagnostic ability of both radiologists and computer-aided diagnosis (CAD) systems. Because of the nonrigidity of the breasts and the 2D projective property of mammograms, this task is not trivial. In this pilot study, we present a computerized framework that differentiates between corresponding images of the same lesion in different views and noncorresponding images, i.e., images of different lesions. A dual-stage segmentation method, which employs an initial radial gradient index (RGI) based segmentation and an active contour model, is applied to extract mass lesions from the surrounding parenchyma. Then various lesion features are automatically extracted from each of the two views of each lesion to quantify the characteristics of density, size, texture and the neighborhood of the lesion, as well as its distance to the nipple. A two-step scheme is employed to estimate the probability that the two lesion images from different mammographic views are of the same physical lesion. In the first step, a correspondence metric for each pairwise feature is estimated by a Bayesian artificial neural network (BANN). Then, these pairwise correspondence metrics are combined using another BANN to yield an overall probability of correspondence. Receiver operating characteristic (ROC) analysis was used to evaluate the performance of the individual features and the selected feature subset in the task of distinguishing corresponding pairs from noncorresponding pairs. Using a FFDM database with 123 corresponding image pairs and 82 noncorresponding pairs, the distance feature yielded an area under the ROC curve (AUC) of 0.81±0.02 with leave-one-out (by physical lesion) evaluation, and the feature metric subset, which included distance, gradient texture, and ROI-based correlation, yielded an AUC of 0.87±0.02. The improvement by using multiple feature metrics was statistically significant compared to single feature performance.
AB - Identifying the corresponding images of a lesion in different views is an essential step in improving the diagnostic ability of both radiologists and computer-aided diagnosis (CAD) systems. Because of the nonrigidity of the breasts and the 2D projective property of mammograms, this task is not trivial. In this pilot study, we present a computerized framework that differentiates between corresponding images of the same lesion in different views and noncorresponding images, i.e., images of different lesions. A dual-stage segmentation method, which employs an initial radial gradient index (RGI) based segmentation and an active contour model, is applied to extract mass lesions from the surrounding parenchyma. Then various lesion features are automatically extracted from each of the two views of each lesion to quantify the characteristics of density, size, texture and the neighborhood of the lesion, as well as its distance to the nipple. A two-step scheme is employed to estimate the probability that the two lesion images from different mammographic views are of the same physical lesion. In the first step, a correspondence metric for each pairwise feature is estimated by a Bayesian artificial neural network (BANN). Then, these pairwise correspondence metrics are combined using another BANN to yield an overall probability of correspondence. Receiver operating characteristic (ROC) analysis was used to evaluate the performance of the individual features and the selected feature subset in the task of distinguishing corresponding pairs from noncorresponding pairs. Using a FFDM database with 123 corresponding image pairs and 82 noncorresponding pairs, the distance feature yielded an area under the ROC curve (AUC) of 0.81±0.02 with leave-one-out (by physical lesion) evaluation, and the feature metric subset, which included distance, gradient texture, and ROI-based correlation, yielded an AUC of 0.87±0.02. The improvement by using multiple feature metrics was statistically significant compared to single feature performance.
KW - Computer-aided diagnosis
KW - Correlative feature analysis
KW - Feature selection
KW - Full-field digital mammography
KW - Lesion segmentation
UR - http://www.scopus.com/inward/record.url?scp=56749169702&partnerID=8YFLogxK
U2 - 10.1118/1.3005641
DO - 10.1118/1.3005641
M3 - Article
C2 - 19175108
AN - SCOPUS:56749169702
SN - 0094-2405
VL - 35
SP - 5490
EP - 5500
JO - Medical Physics
JF - Medical Physics
IS - 12
ER -