“…For example, currently available R and MATLAB statistical software packages for HRV analysis mostly deal with outliers by allowing the user to manipulate directly the HRV records (Perakakis et al, 2010;Rodríguez-Liñares et al, 2011). This kind of pre-processing, however, comes with a major risk: the filtered trajectory might be clear of evident outliers, but other atypical observations incompatible with the assumed model remain hidden within the processed sample.…”