Extracting information from apparent randomness in cardiovascular data