Using Workbench Command

DO TFCE ON A METRIC FILE
   wb_command -metric-tfce
      <surface> - the surface to compute on
      <metric-in> - the metric to run TFCE on
      <metric-out> - output - the output metric

      [-presmooth] - smooth the metric before running TFCE
         <kernel> - the size of the gaussian smoothing kernel in mm, as sigma
            by default

         [-fwhm] - kernel size is FWHM, not sigma

      [-roi] - select a region of interest to run TFCE on
         <roi-metric> - the area to run TFCE on, as a metric

      [-parameters] - set parameters for TFCE integral
         <E> - exponent for cluster area (default 1.0)
         <H> - exponent for threshold value (default 2.0)

      [-column] - select a single column
         <column> - the column number or name

      [-corrected-areas] - vertex areas to use instead of computing them from
         the surface
         <area-metric> - the corrected vertex areas, as a metric

      Threshold-free cluster enhancement is a method to increase the relative
      value of regions that would form clusters in a standard thresholding
      test.  This is accomplished by evaluating the integral of:

      e(h, p)^E * h^H * dh

      at each vertex p, where h ranges from 0 to the maximum value in the data,
      and e(h, p) is the extent of the cluster containing vertex p at threshold
      h.  Negative values are similarly enhanced by negating the data, running
      the same process, and negating the result.

      When using -presmooth with -corrected-areas, note that it is an
      approximate correction within the smoothing algorithm (the TFCE
      correction is exact).  Doing smoothing on individual surfaces before
      averaging/TFCE is preferred, when possible, in order to better tie the
      smoothing kernel size to the original feature size.

      The TFCE method is explained in: Smith SM, Nichols TE., "Threshold-free
      cluster enhancement: addressing problems of smoothing, threshold
      dependence and localisation in cluster inference." Neuroimage. 2009 Jan
      1;44(1):83-98. PMID: 18501637