Sotirios Athanasios Tsaftaris1, Xiangzhi Zhou2, Richard Tang2, Rachel Klein2, Rohan Dharmakumar2
1Electrical Engineering and Computer Science, Northwestern University, Evanston, IL, USA; 2Radiology, Northwestern University, Chicago, IL, USA
A critical component in computing quantitative diagnostic metrics, such as ejection fraction, as well as image segmentation and registration, is the accurate identification of the end-systolic (ES) and end-diastolic (ED) frames in cine MRI. Localization of the LV is also important, to assist further analysis (ie., myocardial segmentation). In this paper we propose an image-driven statistical method that utilizes cross-correlation of pixels, to detect ES and ED images, as well as, localize the LV, from cine MRI acquired from canines under control conditions. The method is fully automated, computationally efficient, and requires no parameterization, initialization, and ROI selection.robust, and can be extended to 4D MRI.