[FieldTrip] Leadfield of one condition with two different 'grad' structures

[Sara Aurtenetxe]

Dear all, 

I would appreciate to get some advice on the following.

I am computing the leadfield (ft_prepare_leadfield) for the subsequent source analysis (ft_sourceanalysis) of ERF data acquired with Elekta Neuromag system. 
My experiment is acquired in a unique run, including 3 randomized conditions for each participant normally. 
Therefore, the unique 'cfg.grad' structure is used for the leadfield computation of each participant.

However, due to a must, the experiment was recorded into two independent runs (splited) for one of the participants. Here comes the issue: 
Given that the 'grad' structure is specific to each run, this participant turns to with two different 'grad' structures for the same condition.

My question is: which is the most suitable approach to compute the leadfield of one condition when the data comes from two different runs, so when having two different 'grad' structures?

I would say that I should compute the leadfield and the sourceanalysis for each run and condition separately, and then join the solutions of the same condition. 
But I wonder if this is a correct approach and/or if there is any additional/more appropriate approach for that. 

Any comment will be very appreciated.

Thank you in advance.
Best,

Sara