In this study, we propose an approach for the assessment of the evolution of heart function in a population of ST-elevation myocardial infarction (STEMI) patients. The patients were imaged using cine MRI with late gadolinium enhancement at 1 month after STEMI and at follow-up 12 months later. We reduced the dimensionality of the temporal strain data in order to represent the segmental strain patterns found in these individuals in a more compact manner. Then, this compact description of strain was used to identify left ventricular segments with abnormal function, as compared to a control group of healthy patients, and to investigate if 1) the compact representation of the patterns is more effective in predicting scarring than traditional measurements; and 2) how the strain patterns evolve over the following months. To that end, we offer a method for tracking strain evolution and, in particular, comparing infarcted regions with remote regions and healthy hearts. On a population of 29 STEMI patients and 18 controls, we found that, despite considering the extent of the lesion, we were not able to identify a clear mechanism of evolution; nonetheless, the technique may be beneficial in subsequent, larger longitudinal studies to quantitatively characterise patient outcomes.