%0 Journal Article %A Jing Tian %A Mark F. Smith %A Ghada Ahmad %A Vasken Dilsizian %A Alejandro Jimenez %A Timm Dickfeld %T Integration of 3-Dimensional Scar Models from SPECT to Guide Ventricular Tachycardia Ablation %D 2012 %R 10.2967/jnumed.111.094904 %J Journal of Nuclear Medicine %P 894-901 %V 53 %N 6 %X The integration of myocardial scar models in 3-dimensional (3D) mapping systems may provide a novel way of helping to guide ventricular tachycardia (VT) ablations. This study assessed the value of 201Tl SPECT perfusion imaging to define ventricular myocardial scar areas and to characterize electrophysiology voltage-derived myocardial substrate categories of scar, border zone (BZ), and normal myocardium regions. Scar and BZ regions have been implicated in the genesis of ventricular arrhythmias. Methods: Ten patients scheduled for VT ablation underwent 201Tl SPECT before the ablation procedure. 3D left ventricular (LV) scar models were created from the SPECT images. These scar models were registered with the LV voltage maps and analyzed with a 17-segment cardiac model. Scar location and scar burden were compared between the SPECT scar models and voltage maps. In addition, 201Tl SPECT uptake was quantified using a 68-segment cardiac model and compared among voltage-defined scar, BZ, and normal segments. Results: 3D models of LV myocardium and scar were successfully created from 201Tl SPECT images and integrated in a clinical mapping system. The surface registration error with the electrophysiology voltage map was 4.4 ± 1.0 mm. The 3D scar location from SPECT matched in 72% of the segments with the voltage map findings. All successful ablation sites were located within the SPECT-defined scar or within 1 cm of its border, with 73% of the successful ablation sites within 1 cm of the scar border. Voltage measurements in SPECT-defined scar and normal areas were 1.2 ± 1.7 and 3.4 ± 2.8 mV, respectively (P < 0.001). The fractional SPECT scar burden area (18.8% ± 5.2%) agreed better with the abnormal (scar plus BZ) voltage area (20.8% ± 15.7%) than with the scar voltage area (5.8% ± 5.8%). Mean normalized 201Tl uptake was 55% ± 21% in the voltage-defined scar, 63% ± 20% in BZ, and 79% ± 17% in normal myocardial segments (P < 0.05 for scar or BZ vs. normal). Conclusion: 3D SPECT surface models of LV scar were accurately integrated into a clinical mapping system and predicted endocardial voltage-defined scar. These preliminary data support the possible use of widely available 201Tl SPECT to facilitate substrate-guided VT ablations. %U https://jnm.snmjournals.org/content/jnumed/53/6/894.full.pdf