Computed Tomography (CT) is widely used for the diagnosis of variety of medical and surgical conditions and acquiring high quality CT diagnostic images with lower radiation would be ideal to reduce the risks of radiation. One method of solving this problem is to focus on optimizing the reconstruction methods in CT. The first version of AREMI uses multi-core GPU for the reconstruction from raw attenuation data measured by a Siemens Definition Flash CT scanner.
Our method provides a better quality reconstruction than the commercial algorithm.
Cardiac magnetic resonance imaging is used to evaluate the right ventricle in a number of congenital heart conditions, such as post-op Tetralogy of Fallot. Currently, the clinical standard is to assess right ventricle systolic function, size, mass, stroke volume and myocardial scarring with MRI. Other parameters such as strain and diastolic function require intense post-processing and are therefore used only in the research setting. We focus on the development of an automated motion tracking program to assess the RV motion.
The advent of real-time 3D echocardiography allows better quantification of size, shape and function of the heart. Although the 3D echocardiography has potential benefits in patient care in a number of ways, including pre- and post-surgical planning, its application is restricted by the limited field-of-view. This limitation can be overcome by fusing multiple images taken from different probe positions. We estimate the probe location using a multi-camera motion tracking system and develop of an image fusion approach for better contrast and field-of-view.
DENSE is a non-invasive MR protocol that provides myocardial displacement mapping with excellent spatial resolution. A cine DENSE sequence provides a series of images over segments of the cardiac cycle, which facilitates analyzing the kinematics over the entire cardiac cycle. This project will assess the potential benefits of using the DENSE sequence in cardiac regional abnormality detection.
The interaction between fluids and structures in the body is vital to the function of many physiological processes, including blood flow in the cardiovascular system. Methods that allow computer simulation of fluid-structure interaction are useful in understanding these interactions. We build fluid-structure interaction simulation models for real patients using 3-D anatomical information from MRI, ultrasound, and CT. These patient-specific simulation models will be used for optimization of surgical and non-invasive interventional procedures.
The assessment of LV is often limited to systolic function and it mainly focuses on the analysis of regional wall motion or ejection fraction. However, recent clinical studies suggest that assessment of diastolic function is as important as systolic function. The diastolic function plays an important role in assessing cardiovascular abnormalities, particularly in the case of heart failure with preserved ejection fraction. We are developing automated methods to assess the LV impaired relaxation using short-axis cine MR images.
With the advent of 3D echocardiography, the functional analysis and visualization of three dimensional data has become increasingly important. The actual heart motion is a complicated combination of motions and it can only be observed accurately in 3D. This project focuses on the development of automatic algorithms for tracking myocardial boundaries, motion estimation and 3D visualization of the 3D echocardiographic data.
