Impact of scan quality on the diagnostic performance of CCTA, SPECT, and PET for diagnosing myocardial ischemia defined by fractional flow reserve Academic Article uri icon

Overview

MeSH Major

  • Diabetes Mellitus, Type 2
  • Social Class
  • Statistics as Topic
  • Women's Health

abstract

  • © 2019 Society of Cardiovascular Computed Tomography Background: Scan quality can have a significant effect on the diagnostic performance of non-invasive imaging techniques. However, the extent of its influence has scarcely been investigated in a head-to-head manner. Methods: Two-hundred and eight patients underwent CCTA, SPECT, and PET prior to invasive fractional flow reserve measurements. Scan quality was classified as either good, moderate, or poor. Results: Distribution of good, moderate, and poor quality scans was; CCTA; 66%, 22%, 13%; SPECT; 52%, 38%, 9%; PET; 86%, 13%, 1%. Good quality CCTA scans possessed a higher specificity (75% vs. 31%, p = 0.001), positive predictive value (PPV, 71% vs. 51%, p = 0.050), and accuracy (80% vs. 60%, p = 0.009) compared to moderate quality scans, while sensitivity (94%) and negative predictive value (NPV, 88%) were similar to moderate and poor quality scans. Sensitivity (76%), NPV (84%), and accuracy (85%) of good quality SPECT scans was superior to those of moderate (41% p = 0.001, 56% p = 0.010, 70% p = 0.010) and poor quality (30% p = 0.003, 65% p = 0.069, 63% p = 0.038). Specificity (92%) and PPV (87%) of good quality SPECT scans did not differ from scans of diminished quality. Good quality PET scans exhibited high sensitivity (84%), specificity (86%), NPV (88%), PPV (81%) and accuracy (85%), which was comparable to scans of lesser quality. Good quality CCTA, SPECT, and PET scans demonstrated a similar diagnostic accuracy (p = 0.247). Conclusion: Diagnostic performance of CCTA, and SPECT is hampered by scan quality, while the diagnostic value of PET is not affected. Good quality CCTA, SPECT, and PET scans possess a high diagnostic accuracy.

publication date

  • January 2019

Research

keywords

  • Academic Article

Identity

Digital Object Identifier (DOI)

  • 10.1016/j.jcct.2019.06.007