--- a
+++ b/notebooks/split_checks.py
@@ -0,0 +1,71 @@
+# %%
+import os
+
+# enable lib loading even if not installed as a pip package or in PYTHONPATH
+from pathlib import Path
+os.chdir(Path(__file__).resolve().parent.parent)
+
+from adpkd_segmentation.data.data_utils import get_labeled, make_dcmdicts  # noqa
+from adpkd_segmentation.datasets.splits import GenSplit  # noqa
+
+# %%
+dcm_paths = sorted(get_labeled())
+dcm2attribs, patient2dcm = make_dcmdicts(tuple(dcm_paths))
+all_patient_IDS = list(patient2dcm.keys())
+
+# %%
+seed = 1
+splitter = GenSplit(seed=seed)
+split = splitter(all_patient_IDS)
+# %%
+print(split["val"])
+
+# %%
+for patient in split["val"]:
+    sequences = set()
+    mrs = set()
+    for dcm in patient2dcm[patient]:
+        sequences.add(dcm2attribs[dcm]["seq"])
+        mrs.add(dcm2attribs[dcm]["MR"])
+    print(sequences)
+    print(mrs)
+
+
+# %%
+print(split["test"])
+
+# %%
+for patient in split["test"]:
+    sequences = set()
+    mrs = set()
+    for dcm in patient2dcm[patient]:
+        sequences.add(dcm2attribs[dcm]["seq"])
+        mrs.add(dcm2attribs[dcm]["MR"])
+    print(sequences)
+    print(mrs)
+
+
+# %%
+print(split["train"])
+
+# %%
+for patient in split["train"]:
+    sequences = set()
+    mrs = set()
+    for dcm in patient2dcm[patient]:
+        sequences.add(dcm2attribs[dcm]["seq"])
+        mrs.add(dcm2attribs[dcm]["MR"])
+    print(sequences)
+    print(mrs)
+
+# %%
+for patient in all_patient_IDS:
+    sequences = set()
+    mrs = set()
+    for dcm in patient2dcm[patient]:
+        sequences.add(dcm2attribs[dcm]["seq"])
+        mrs.add(dcm2attribs[dcm]["MR"])
+    print(sequences)
+    print(mrs)
+
+# %%