DETECTING A CONDITION OF A PATIENT USING A PROBABILITY-CORRELATION BASED MODEL

Techniques are described for detecting a condition of a patient using a probability-correlation based model that integrates a plurality of parameters associated with the condition. A medical device that operates in accordance with the techniques obtains a plurality of parameters associated with the condition of the patient. The medical device obtains probabilities that the condition of the patient exists based on each single parameter separately and correlations between each of the parameters and the other ones of the parameters. After obtaining the probabilities and correlations associated with each of the parameters, the medical device determines whether the condition of the patient exists based on the determined probabilities and correlations. Such techniques may be particularly effective for use in distinguishing whether a rhythm of a patient is treatable, e.g., VT or VF, or non-treatable, e.g., SVT.