Ischemic events are associated with an increased dispersion of ventricular repolarization, and its related ventricular arrhythmia risk. Time-warping-based shape markers 〖(d〗w) have been incorporated into T wave repolarization analysis trying to capture global morphological variations. This work proposes to use the dw marker, estimated only in the T-wave peak-to-end interval, to capture ischemia-induced dispersion of repolarization. Restricting the analysis to the later area of the T wave avoids the influence of ST segment elevation/depression in the early part of the T wave. ECG from the STAFF-III database during elective balloon percutaneous coronary intervention (PCI), lasting for an average of 4 minutes, were analyzed, together with their baseline recordings. Spatial Principal Components Analysis were used to generate a transformed lead emphasizing T-wave. A methodology for continuous Mean Warped T-Wave (MWTW) estimation along the recordings together with a reference MWTW election criteria was developed to compute the d_w series. A normalized marker, R_d^PCA , comparing d_w changes during PCI relative to those at control recordings, was proposed, providing information on the magnitude of change generated by ischemia. T-wave changes during PCI, were accompanied by d_w changes with average value at the end of the occlusion of 7.38 ms (standard deviation: 5.25), and following a gradually increasing trend as inflation time progresses. On the contrary, d_w during control recordings remains stable with average value of 1.07 ms (standard deviation: 0.21). In average, 24% of the maximum d_w measured at PCI completion was reached after the first third of the occlusion. After the second third, 40% of the maximum d_w was measured. Repolarization changes appeared during the PCI procedure in the total study population, R_d^PCA >1 in all the studied patients. The T wave time-warping shape marker, d_w, based on the late T wave area allows to monitor ischemia-induced repolarization dispersions changes.