add neuromelanin column to manifest

neurodatascience · Aug 30, 2024 · 51bc2a4 · 51bc2a4
1 parent 752dc9e
commit 51bc2a4
Showing 1 changed file with 10 additions and 3 deletions.
diff --git a/nipoppy/workflow/tabular/generate_manifest.py b/nipoppy/workflow/tabular/generate_manifest.py
@@ -3,6 +3,7 @@
 import argparse
 import datetime
 import json
+import re
 import shutil
 import warnings
 from pathlib import Path
@@ -57,6 +58,10 @@
 GLOBAL_CONFIG_VISITS = 'VISITS'
 GLOBAL_CONFIG_TABULAR = 'TABULAR'
 
+COL_NEUROMELANIN_MANIFEST = 'neuromelanin'
+RE_NEUROMELANIN = '[nN][mM]' # neuromelanin pattern
+COLS_MANIFEST = COLS_MANIFEST + [COL_NEUROMELANIN_MANIFEST]
+
 DATATYPES = [DATATYPE_ANAT, DATATYPE_DWI, DATATYPE_FUNC]
 
 # flags
@@ -218,9 +223,11 @@ def run(global_config_file: str, regenerate: bool, make_release: bool):
 
     # create imaging datatype availability lists
     seen_datatypes = set()
-    df_imaging = df_imaging.groupby([COL_SUBJECT_MANIFEST, COL_VISIT_MANIFEST, COL_SESSION_MANIFEST])[COL_DATATYPE_MANIFEST].aggregate(
-        lambda descriptions: get_datatype_list(descriptions, description_datatype_map, seen=seen_datatypes)
-    )
+    df_imaging[COL_NEUROMELANIN_MANIFEST] = df_imaging[COL_DATATYPE_MANIFEST]
+    df_imaging = df_imaging.groupby([COL_SUBJECT_MANIFEST, COL_VISIT_MANIFEST, COL_SESSION_MANIFEST])[[COL_DATATYPE_MANIFEST, COL_NEUROMELANIN_MANIFEST]].aggregate({
+        COL_DATATYPE_MANIFEST: lambda descriptions: get_datatype_list(descriptions, description_datatype_map, seen=seen_datatypes),
+        COL_NEUROMELANIN_MANIFEST: lambda descriptions: any([re.search(RE_NEUROMELANIN, description) for description in descriptions]),
+    })
     df_imaging = df_imaging.reset_index()
     print(f'\nFinal imaging dataframe shape: {df_imaging.shape}')