The objective of this research is to improve the measurement of coronary artery lesion dimensions from x-ray cine angiograms (""""""""Quantitative Coronary Angiography"""""""") by optimally utilizing the information available in the entire multiple-image sequence. Quantitative coronary angiography is an important endpoint for clinical investigations of new drug, device, diet and lifestyle interventions for coronary artery disease. Conventional quantitative coronary angiography derives measurements can achieve a precision of 0.15 mm or better under ideal conditions, actual precision is often far worse due to inherent sources of noise, such as overlapping patient structures and x-ray quantum mottle.
We aim to develop and validate an automated method for performing these measurements in three potential benefits: l)Increasing the number of independent measurements should reduce the effect of noise and thereby improve precision. 2)Multiple sequential frames can provide qualitatively different data not available in single frames, such as coronary artery velocity, which can be used to improve accuracy by detecting and correcting for motion blur, or by improving segmentation of the artery of interest from other structures with different motions. 3)Measuring dimensions throughout the cardiac cycle can provide dynamic information related to the elastic properties of the coronary artery wall, which are known to be affected by atherosclerosis and vasoactive drugs. Our research plan has three sections, reflecting our three specific aims. In A we aim to develop optimal methods for automatically tracking the artery and for utilizing interframe motion to separate the artery from background structures. In B we aim to develop methods for automated measurement of coronary lesion diameter based on detection of arterial edges in multiple sequential frames. In C we aim to develop methods for densitometric measurement of coronary artery cross-sectional area using improved background and scatter corrections based on the information in multiple and sequential frames. We will evaluate these new analysis methods in animal and archived clinical coronary angiograms compare the results with conventional single frame methods.
| Close, Robert A; Abbey, Craig K; Morioka, Craig A et al. (2002) Evaluation of layer decomposition for multiframe quantitative coronary angiography. Med Phys 29:311-8 |
| Close, Robert A; Abbey, Craig K; Whiting, James S (2002) Improved localization of coronary stents using layer decomposition. Comput Aided Surg 7:84-9 |
| Close, R A; Abbey, C K; Morioka, C A et al. (2001) Accuracy assessment of layer decomposition using simulated angiographic image sequences. IEEE Trans Med Imaging 20:990-8 |
| Abbey, C K; Eckstein, M P (2000) Derivation of a detectability index for correlated responses in multiple-alternative forced-choice experiments. J Opt Soc Am A Opt Image Sci Vis 17:2101-4 |
| Morioka, C A; Abbey, C K; Eckstein, M et al. (2000) Simulating coronary arteries in x-ray angiograms. Med Phys 27:2438-44 |
| Close, R A; Shah, K C; Whiting, J S (1999) Regularization method for scatter-glare correction in fluoroscopic images. Med Phys 26:1794-801 |
