? Cardiac perfusion imaging by single-photon emission computed tomography (SPECT) is the most frequently performed study in clinical nuclear medicine, where it plays an important role in detection and evaluation of coronary artery disease. In clinical practice, the image frames in a gated SPECT study are usually computed separately, one by one, by standard image-reconstruction methods. Because individual frames of gated SPECT data can be very noisy, this frame-by-frame reconstruction approach yields relatively poor image quality, which can limit the effectiveness of gated SPECT for assessing LV function. In the prior project period, we developed new, four-dimensional (4D) image-reconstruction techniques, which can significantly improve the quality of gated SPECT images. In these techniques, we consider the entire image sequence as a single spatio-temporal signal, which is 4D (defined by three spatial dimensions, plus time). By this approach, we exploit the fact that the desired signal in gated SPECT is slowly varying in time, whereas the noise is uncorrelated from frame to frame. This grant application is a competing continuation proposal to conduct research that will have two main components. In the first component, we will complete our research on 4D image reconstruction for gated SPECT by refining the required cardiac motion modeling and performing clinical evaluations of the best candidate methods. In the second component, we will develop a new type of SPECT imaging study we have proposed, called 5D SPECT, which unifies gated SPECT and dynamic SPECT into a single imaging method. By using list-mode data, 5D imaging can produce an image sequence that shows both time-varying cardiac motion and tracer distribution. In addition to providing the usual benefits of clinical, gated SPECT imaging, 5D SPECT may: 1) permit visualization of time variations in cardiac function during recovery from stress (e.g., wall motion, ejection fraction, wall thickening, and tracer kinetics); 2) produce more-accurate images by accounting for """"""""upward creep""""""""; and 3) add pertinent information about tracer kinetics to that provided by traditional gated SPECT. By using gated information, 5D imaging may also improve upon dynamic SPECT by reducing kinetic-modeling errors caused by cardiac motion.
The specific aims will be: 1) to refine our 4D techniques, and prepare for 5D research, by incorporating more-accurate motion-modeling and estimation methods; 2) to develop 5D imagereconstruction techniques for gated/dynamic SPECT imaging, including: (a) separable 5D methods; and (b) fully 5D methods; and 3) to develop and conduct task-based performance evaluations of the methods developed in Aims 1 and 2. ? ? ?

National Institute of Health (NIH)
National Heart, Lung, and Blood Institute (NHLBI)
Research Project (R01)
Project #
Application #
Study Section
Biomedical Imaging Technology Study Section (BMIT)
Program Officer
Pandit, Sunil
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
Illinois Institute of Technology
Engineering (All Types)
Schools of Engineering
United States
Zip Code
Marin, Thibault; Kalayeh, Mahdi M; Parages, Felipe M et al. (2014) Numerical Surrogates for Human Observers in Myocardial Motion Evaluation From SPECT Images. IEEE Trans Med Imaging 33:38-47
Kalayeh, Mahdi M; Marin, Thibault; Brankov, Jovan G (2013) Generalization Evaluation of Machine Learning Numerical Observers for Image Quality Assessment. IEEE Trans Nucl Sci 60:1609-1618
Brankov, Jovan G (2013) Evaluation of the channelized Hotelling observer with an internal-noise model in a train-test paradigm for cardiac SPECT defect detection. Phys Med Biol 58:7159-82
Niu, Xiaofeng; Yang, Yongyi; King, Michael A (2012) Comparison study of temporal regularization methods for fully 5D reconstruction of cardiac gated dynamic SPECT. Phys Med Biol 57:5523-42
Qi, Wenyuan; Yang, Yongyi; Niu, Xiaofeng et al. (2012) A quantitative study of motion estimation methods on 4D cardiac gated SPECT reconstruction. Med Phys 39:5182-93
Niu, Xiaofeng; Yang, Yongyi (2011) Tomographic reconstruction of gated data acquisition using DFT basis functions. IEEE Trans Image Process 20:176-85
Massanes, Francesc; Cadennes, Marie; Brankov, Jovan G (2011) Compute-unified device architecture implementation of a block-matching algorithm for multiple graphical processing unit cards. J Electron Imaging 20:
Niu, Xiaofeng; Yang, Yongyi; Jin, Mingwu et al. (2011) Effects of motion, attenuation, and scatter corrections on gated cardiac SPECT reconstruction. Med Phys 38:6571-84
Marin, Thibault; Brankov, Jovan G (2010) Deformable left-ventricle mesh model for motion-compensated filtering in cardiac gated SPECT. Med Phys 37:5471-81
Li, Ling; Yang, Yongyi (2010) Optical flow estimation for a periodic image sequence. IEEE Trans Image Process 19:1-10

Showing the most recent 10 out of 20 publications