This paper addresses the effect of isolation on the spread of Covid-19 in the State of Minas Gerais, Brazil. The compartmental model SEIR (Susceptible, Exposed, Infectious, Recovered) has been employed. The parameters of the model have been estimated with the help of optimization techniques based on evolutionary strategies. From the model obtained, this work evaluates the impacts of reducing the isolation in the peak of infected people. Results show that the isolation has reduced the peak in 94%, which has played a crucial role in avoiding the health system collapse up to now. Moreover, around 70% of isolation guarantees a good quality of health system service to deal with the pandemic of Covid-19 according to the number of intensive care units and considering a reproductive number equals to 3.60. This is an important result to be considered during easing physical distancing measures.