Ocular blood flow as a clinical observation: Value, limitations and data analysis