The assessment of autonomic activity in various clinical contexts relies on the analysis of heart rate variability (HRV). Standard HRV analyses often fall short in effectively disentangling sympathetic and parasympathetic activations. Recently, we introduced an HRV methodology based on Poincaré plot analysis. This approach provides time-resolved estimations of the primary HRV rhythms, separately from the changes in the baseline interbeat interval duration. It utilizes a robust covariance matrix estimation technique grounded in the Ledoit-Wolf approach. Through this method, we derived the Cardiac Sympathetic and Parasympathetic Indices (CSI, CPI) and compared them to the pre and post freezing of gait (FOG) episodes in patients with Parkinson's disease. Our method outperforms standard spectral counterparts of HRV in monitoring autonomic activity. This highlights their value for future development in wearables for real-time monitoring of FOG episodes through HRV analysis.