Toward feedback temperature control for retinal laser treatment

Automatic feedback temperature control has been proposed recently as a promising approach for retinal thermal laser therapy. It can regulate the retinal temperature increase to a pre-specified desired value, which can provide homogeneous temperature increase at the different irradiation sites as the feedback control can automatically set the required corresponding laser power. Therefore, we can avoid over treatment and its consequences due to the manual setting often provided by ophthalmologists in practice. However, in order to achieve successful automatic treatment, a mathematical model for the process is necessary to synthesize the feedback control algorithm. In this study, we develop a model, which can describe the dynamical relation between the temperature increase and the applied laser power at the different irradiation sites. We adopt the system identification approach, which is used to build mathematical models for dynamical systems based on experimental measurements. The identified model here can fit the rear-time data with an accuracy of about 91 ± 1.91%, which indicates its validity for effective feedback control. The control algorithm based on such models can achieve consistent irradiation time of about 50 ms in simulation