Your organization runs a backend service on Compute Engine instances that occasionally experience memory usage spikes exceeding 75%. The memory usage typically drops below 60% within a few minutes. You have set up an alert to trigger whenever memory usage exceeds 75%, but this is resulting in frequent false positives. You want to configure the alert to notify you only when memory usage remains above 75% for at least 7 minutes. How should you achieve this?