Volumetric Quantification of Myocardial Perfusion Using Analysis of Multi-Detector Computed Tomography 3D Datasets: Comparison With Nuclear Perfusion Imaging

OBJECTIVES: Although the ability of multi-detector computed tomography (MDCT) to detect perfusion abnormalities associated with acute and chronic myocardial infarction (MI) has been demonstrated, this methodology is based on visual interpretation of selected 2D slices. We sought to develop a new technique for quantitative volumetric analysis of myocardial perfusion from 3D datasets and test it against resting nuclear myocardial perfusion imaging (NMPI) reference.

METHODS: We studied 44 patients undergoing CTCA: a control group of 15 patients and a study group of 29 patients. MDCT datasets acquired for CTCA were analyzed using custom software designed to: (1) generate bull's eye display of myocardial perfusion and (2) calculate a quantitative index of extent and severity of perfusion abnormality, Q(H), for 16 volumetric myocardial segments. Visual interpretation of MDCT-derived bull's eyes was compared with rest NMPI scores using kappa statistics of agreement on a coronary territory and patient basis. Quantitative MDCT perfusion data were correlated with rest NMPI summed scores and used for objective detection of perfusion defects.

RESULTS: Visual analysis of MDCT-derived bull's eyes accurately detected perfusion defects in agreement with NMPI (kappa = 0.70 by territory; 0.79 by patient). Quantitative data were in good agreement with NMPI, as reflected by: (1) correlation of 0.87 (territory) and 0.84 (patient) between summed Q(H) and NMPI scores, (2) area under ROC curve 0.87 with sensitivity of 0.79-0.92, specificity 0.83-0.91, and accuracy 0.83-0.89 for objective detection of abnormalities.

CONCLUSIONS: Our new technique for volumetric analysis of 3D MDCT images allows accurate objective detection of perfusion defects. This perfusion information can be obtained without additional radiation or contrast load, and may aid in elucidating the significance of coronary lesions.



Posted in Computed Tomography and tagged , , , , , .


  1. An interesting method for the (more objective) assessment of myocardial perfusion defects with coronary CTA. The study compared findings with SPECT.

  2. These software programs are important to further evaluate the clinical utility of cardiac CT. Another example is atherosclerosis imaging (see below paper by Dey et al.)
    However, the ability to assess these characteristics more objectively, or the implementation on commercial systems; does not imply clinical utility, which needs to be demonstrated in studies with clinical endpoints (MACE).

    Automated 3-dimensional quantification of noncalcified and calcified coronary plaque from coronary CT angiography.
    Damini Dey, Victor Y. Cheng, Piotr J. Slomka, Ryo Nakazato, Amit Ramesh, Swaminatha Gurudevan, Guido Germano, Daniel S. Berman.
    Journal of Cardiovascular Computed Tomography, October 2009, pages 372-382.
    PMID: 20083056

  3. One thing that I have noticed while doing coronary CTA examinations and by over-reading the thoracic part of the study for my cardiology colleagues, is that an under-perfused segment of the LV can have more than 1 meaning. In addition to the potential of this being a perfusion defect revealing an acute process, it may also represent a foci of fibrous replacement from a remote chronic infarct. Carefully reviewing non-enhanced images of the chest (if available) and looking for these areas are of utmost importance.
    Is anyone aware of studies differentiating these 2 types of perfusion defects based on attenuation values?

  4. That’s a very good question! I wonder whether it is possible to define newer and older perfusion defects by CT since both represent reduced microvascular blood flow. However, it sounds logical as the standard definition for infarct by SPECT using automated quantitated reading is 60% of peak counts. This is related to the degree of reduced blood flow in infarcted myocardium vs. ischemic myocardium.

    Perhaps the answer can be found during stress-CT?

Comments are closed.