Summary
Aim: The simultaneous computation and display of wall motion and perfusion patterns in a single 3D ventricular model would considerably ease the assessment of ECG-gated Tc-99m-sestamibi SPECT, yet the effect on the accuracy of allocating regional perfusion has so far not been validated. Methods: 3D perfusion mapping (3D Perfusion/Motion Map Software) was compared to the visual assessment of ungated tomographic slices and polar perfusion mapping (Cedars-Sinai PTQ) by correlation analysis and receiver operating characteristics (ROC) analysis at different cut-off levels for coronary stenoses in 50 patients (11 single-, 22 two-, 16 three-vessel disease). Ungated SPECT data were obtained by adding the intervals prior to reconstruction and displaying conventional tomographic slices. All display options were visually assessed in 8 ventricular segments according to a 4-point scoring system and compared to the graded results of coronary angiography. Results: All three display options showed a comparable diagnostic performance for the detection of severe stenoses. The diagnostic gain for the detection of stenoses above 59% was highest for ungated tomographic slices, followed by ungated polar mapping and 3D mapping. Regional assessment revealed a limited performance of 3D mapping in the proximal anterior and distal lateral wall. Polar mapping showed a balanced regional performance. Conclusion: 3D Perfusion mapping provides comparable information to conventional display options with the highest diagnostic strength in severe stenoses. Further improvement of the algorithm is needed in the definition of the valve plane.
Urban positioning: 3D mapping‐aided GNSS using dual‐frequency pseudorange measurements from smartphones
