A new method for myocardial ischemia detection is proposed in this communication. The originality of this method relies on the analysis of the dynamics of times series extracted from the ECG, whereas traditional methods are based on static decision rules. After the extraction of a feature vector, from ECG signals from the STAFF3 database, the dynamics are characterised with an Hidden Semi-Markovian Model (HSMM). The ischemic detector uses a reference HSMM and an ischemic HSMM and then compare the log-likelihood of the time series. Results obtained with percutaneous transluminal coronary angioplasty (PTCA) records of the STAFF3 database show an improved detection rate (96% of sensibility and 80% of specificity) with respect to other methods applied on the same database.