search

finnlennartsson / LIM

Less is More
0 stars 0 forks source link

Conversion of T2 M-CRIB from GBG - generating too many files ("run-001 + run-002") #5

Open finnlennartsson opened 1 year ago

finnlennartsson commented 1 year ago

The current conversion of GBG data into BIDS is using heudiconv and the gbg_lim_heuristic.py https://github.com/finnlennartsson/LIM/blob/76049af401fb32359d394afa8e92484ea70cdfdd/bids/heudiconv_heuristics/gbg_lim_heuristic.py#L68-L70 However, this is giving multiple runs (run-001 and run-002) from the same series.

The reason is probably because both normalised and the original series is converted (normalised => run-001 and original => run-002). If we have a re-run, e.g. in 2006, we get four files run-001+run-002 and run-003+run-004).

For LU data, we take "normalised" images https://github.com/finnlennartsson/LIM/blob/76049af401fb32359d394afa8e92484ea70cdfdd/bids/heudiconv_heuristics/lim_heuristic.py#L67-L69

The same should be for GBG data.

So we should change the IF-statement in https://github.com/finnlennartsson/LIM/blob/76049af401fb32359d394afa8e92484ea70cdfdd/bids/heudiconv_heuristics/gbg_lim_heuristic.py#L68