Provision of decision support for acute myocardial infarction

A method of providing decision support for early assessment of patients with suspected acute myocardial infarction (AMI), comprising the steps of:

• measuring a set of n variables related to AMI of a patient at at least one measuring time instance, where n≧2;
• generating a classification of said patient by means of at least one artificial neural network, trained on a training set of n measured variables from m classification groups of typical patients, where m≧2;
• and
• where at least one of said variables related to AMI is a derived quantity from intermittent/continuous ECG/VCG.

A method of providing decision support for early assessment of patients with suspected acute myocardial infarction (AMI), comprising the steps of:

• measuring a set of n variables related to AMI of a patient at at least one measuring time instance, where n≧2;
• generating a classification of said patient by means of at least one artificial neural network, trained on a training set of n measured variables from classification groups of typical patients;
• and
• tuning the performance of said artificial neural network/networks to clinical requirements on predictive values of artificial neural network output in given prevalence situations.

A device for providing decision support for early assessment of patients with suspected acute myocardial infarction (AMI), comprising:

• at least one artificial neural network, trained on a training set of n measured variables related to AMI from m classification groups of typical patients, and arranged for classifying said patient, where n, m≧2;
• means for inputting a set of said n measured values related to AMI of a patient from at least one measuring time instance;
• and
• where at least one of said variables related to AMI is a derived quantity from intermittent/continuous ECG/VCG.

A device for providing decision support for early assessment of patients with suspected acute myocardial infarction (AMI), comprising:

• at least one artificial neural network, trained on a training set of n measured variables related to AMI from m classification groups of typical patients, and arranged for classifying said patient, where n, m≧2;
• means for inputting a present set of said n measured values related to AMI of a patient from at least one measuring time instance to generate a classification of said patient;
• and
• means for tuning the performance of said artificial neural network/networks to clinical requirements on predictive values of output in given prevalence situations.