Because from my reading that's not the number of hours they used the machines, but the number of hours in which the machines were available to be used. And since they also state that their utilization figures only include emergency cases (and exclude scheduled scans) I don't consider it a particularly relevant source in the first place.