A system and method for detecting a drivers sudden heart attack applies to detecting whether the driver of a vehicle is suffering a sudden heart attack. The method comprises steps: persistently capturing vital-sign signals of respiration rate, heart rhythm and blood pressure of the driver and transmitting the vital-sign signals to a monitoring system a processor of the monitoring system using an artificial neural network to train the vital-sign signals and establish personalized models specific to the driver and respectively having threshold values the monitoring system examining whether any one of the vital-sign signals exceeds its threshold value if at least one type of the vital-sign signals exceeds the threshold value thereof, determining a risk level of the driver according to the quantity of the types of the vital-sign signals exceeding the threshold values thereof, and emitting an alert if necessary.
