Objectives: To demonstrate the feasibility of contrast-enhanced dual-energy mammography (CEDEM) using titanium (Ti) filtering at 49 kVp for high-energy images and a novel artefact reducing image-subtraction post-processing algorithm.

Methods: Fifteen patients with suspicious findings (ACR BI-RADS 4 and 5) detected with digital mammography (MG) that required biopsy were included. CEDEM examinations were performed on a modified prototype machine. Acquired HE and low-energy raw data images were registered non-rigidly to compensate for possible subtle tissue motion. Subtracted CEDEM images were generated via weighted subtraction, using a fully automatic, locally adjusted tissue thickness-dependent subtraction factor to avoid over-subtraction at the breast border. Two observers evaluated the MG and CEDEM images according to ACR BI-RADS in two reading sessions. Results were correlated with histopathology.

Results: Seven patients with benign and eight with malignant findings were included. All malignant lesions showed a strong contrast enhancement. BI-RADS assessment was altered in 66.6 % through the addition of CEDEM, resulting in increased overall accuracy. With CEDEM, additional lesions were depicted and false-positive rate was reduced compared to MG.

Conclusions: CEDEM using Ti filtering with 49 kVp for HE exposures is feasible in a clinical setting. The proposed image-processing algorithm has the potential to reduce artefacts and improve CEDEM images.

Key Points: • CEDEM with a titanium filter is feasible in a clinical setting. • Breast thickness-dependent image subtraction has the potential to improve CEDEM images. • The proposed image-processing algorithm reduces artefacts.

Download full-text PDF

Source
http://dx.doi.org/10.1007/s00330-015-4007-6DOI Listing

Publication Analysis

Top Keywords

cedem images
12
cedem
10
cedem titanium
8
filtering kvp
8
acr bi-rads
8
feasible clinical
8
clinical setting
8
proposed image-processing
8
image-processing algorithm
8
improve cedem
8

Similar Publications

Want AI Summaries of new PubMed Abstracts delivered to your In-box?

Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!