Automatically detected electrocardiogram (ECG) features are used to calculate risk parameters for patients with cardiovascular diseases. In this context, mice are often used for experiments in terms of heart disease models. Therefore, the analysis of murine ECG signals is a topic of great interest for preclinical research. In particular, automatic methods with little or no manual intervention are desirable. However, despite the high number of similarities between humans and mice, there are two major differences. On the one hand, the heart rates of mice are multiple times higher than human heart rates. And on the other hand, the different shapes of the action potentials and their consequences for the ECG morphology make it difficult to use human ECG analysis algorithms for murine data. The aim of this master thesis was the implementation, evaluation, description, and application of an algorithm for automatic feature annotation of murine ECG signals. The algorithm is based on the AIT ECGsolver, which is used for the ECG analysis for human data. The problem of the different ECG morphology is addressed by the implementation of a new feature detection for the QRS-offset and the T-wave features. The developed murine algorithm performs better than the human algorithm. Both the detection rates and the differences from the detected feature to their target values provided by the manual annotations have been improved. A sensitivity up to 91.6 % and a positive prediction up to 94.18 % can be achieved by different algorithm versions.