Chagas disease, endemic to Latin America and increasingly present in non-endemic regions due to migratory flows, often progresses to cardiac involvement with significant disruptions in the electrical conduction system. These alterations can be identified through electrocardiogram (ECG) recordings. However, there is no clear consensus on the criteria that relate ECG changes with the disease course. This work assesses different signal processing techniques to quantify ECG alterations associated with Chagas disease. The analysis is based on a database of clinically classified high-resolution ECGs as normal or abnormal, where the latter case corresponds exclusively to seropositive individuals, and the former case includes both seronegative and seropositive subjects. Temporal analysis involves autoregressive modeling of ECG segments and complexity estimation of heart rate variability (HRV) using five entropy metrics. Frequency analysis is based on fitting the HRV spectrum to a power-law function. Similarities in the resulting feature spaces are evaluated using multidimensional scaling. Our results show that among the signal processing methods, power-law parameters yield the clearest separation between normal and abnormal ECGs, whereas the other techniques reveal partially overlapping patterns. These findings highlight potential quantitative markers for the objective assessment of ECG alterations. Moreover, our approach, based on low-dimensional representations, proves useful for interpreting ECG data in the context of Chagas disease.