How do I develop a Cox proportional hazards model with natural logarithm time-dependence?

I have been using R to evaluate my data; however, it is more advantageous to use MATLAB at the moment. When judging the scaled Schoenfeld residuals against ln(time) I have seen non-proportional hazards with linearity with ln(time). The 'survival' package of R offered to add a time-dependent term to certain covariates giving the following coefficent:

.

I want to add this time dependence to one of the covariates in my model how can I achieve this in MATLAB? If you need further explanation the procedure that I wish to follow is given in the Continuous time-dependent coefficients section on page 20 in the following pdf.

https://cran.r-project.org/web/packages/survival/vignettes/timedep.pdf

Thank you!