Perform QC for LR stacks from NeSVoR using various statistics (image …

…sigma, sigma_var and var)

fetal_brain_utils/cli/compute_nesvor_uq_lr.py

@@ -0,0 +1,152 @@
+from pathlib import Path
+from bids.layout import BIDSLayout
+import argparse
+import re
+import pandas as pd
+import nibabel as ni
+import numpy as np
+
+
+def iter_bids_custom(
+    bids_layout,
+    extension="nii.gz",
+    datatype="anat",
+    suffix="sigma",
+    return_type="filename",
+):
+    for sub in sorted(bids_layout.get_subjects()):
+        for ses in [None] + sorted(bids_layout.get_sessions(subject=sub)):
+            use_suffix = suffix if suffix != "sigma_var" else "var"
+
+            out = bids_layout.get(
+                subject=sub,
+                session=ses,
+                extension=extension,
+                datatype=datatype,
+                suffix=use_suffix,
+                return_type=return_type,
+            )
+            if suffix == "sigma_var":
+                out = [o for o in out if "sigma" in o]
+            yield (sub, ses, out)
+
+
+def main(argv=None):
+    p = argparse.ArgumentParser(
+        formatter_class=argparse.ArgumentDefaultsHelpFormatter,
+        description="Generate measures of uncertainty for the LR inputs given to NeSVoR.",
+    )
+    p.add_argument(
+        "--data_path",
+        default=None,
+        required=True,
+        help="Path to the data.",
+    )
+
+    p.add_argument(
+        "--out_csv",
+        default=None,
+        required=True,
+        help="Out path for the data.",
+    )
+
+    args = p.parse_args(argv)
+
+    data_path = Path(args.data_path).resolve()
+    out_csv = Path(args.out_csv).resolve()
+
+    # Load a dictionary of subject-session-paths
+    # sub_ses_dict = iter_dir(data_path, add_run_only=True)
+    bids_layout = BIDSLayout(data_path, validate=False)
+
+    # Create a variety of measurements based on the
+    # sigma, var, and sigma_var maps.
+    # They can be aggregated based on their mean, median, variance or
+    # mad (median absolute deviation)
+    df = pd.DataFrame(
+        columns=[
+            "sub",
+            "ses",
+            "run",
+            "run_rec",
+            "sigma_avg",
+            "var_avg",
+            "sigma_var_avg",
+            "n_qc",
+            "sigma_med",
+            "var_med",
+            "sigma_var_med",
+            "sigma__var",
+            "var__var",
+            "sigma_var__var",
+            "sigma_mad",
+            "var_mad",
+            "sigma_var_mad",
+        ]
+    )
+
+    for sub, ses, out in iter_bids_custom(bids_layout, suffix="sigma"):
+        for rec_run in [0, 1]:
+            out_rec = [o for o in out if f"slices_run-{rec_run}" in o]
+            if len(out_rec) == 0:
+                continue
+            ids = [int(re.search(r"_run-(\d+)_", element).group(1)) for element in out_rec]
+
+            for i in range(max(ids)):
+                if i not in ids:
+                    continue
+
+                out_id = [el for id_, el in zip(ids, out_rec) if id_ == i]
+                sigma_tot = []
+                var_tot = []
+                sigma_var_tot = []
+                for el in out_id:
+                    slice_sigma = ni.load(el).get_fdata()
+                    slice_T2w = ni.load(el.replace("_sigma", "_T2w")).get_fdata()
+                    slice_var = ni.load(el.replace("_sigma", "_var")).get_fdata()
+                    slice_sigma_var = ni.load(el.replace("_sigma", "_sigma_var")).get_fdata()
+                    loc = slice_T2w > 0
+                    sigma = slice_sigma[loc]
+                    var = slice_var[loc]
+                    sigma_var = slice_sigma_var[loc]
+
+                    if len(sigma) == 0:
+                        print("Empty slice ", Path(el).name)
+                        continue
+                    sigma_tot += sigma.tolist()
+                    var_tot += var.tolist()
+                    sigma_var_tot += sigma_var.tolist()
+
+                def mad(x):
+                    return np.median(np.abs(x - np.median(x)))
+
+                d = {
+                    "sub": sub,
+                    "ses": ses,
+                    "run": i,
+                    "run_rec": rec_run,
+                    "sigma_avg": np.mean(sigma_tot),
+                    "var_avg": np.mean(var_tot),
+                    "sigma_var_avg": np.mean(sigma_var_tot),
+                    "n_qc": len(sigma_tot),
+                    "sigma_med": np.median(sigma_tot),
+                    "var_med": np.median(var_tot),
+                    "sigma_var_med": np.median(sigma_var_tot),
+                    "sigma__var": np.var(sigma_tot),
+                    "var__var": np.var(var_tot),
+                    "sigma_var__var": np.var(sigma_var_tot),
+                    "sigma_mad": mad(sigma_tot),
+                    "var_mad": mad(var_tot),
+                    "sigma_var_mad": mad(sigma_var_tot),
+                }
+                print(d)
+                # Add d to the dataframe:
+                df = pd.concat([df, pd.DataFrame(d, index=[0])], ignore_index=True)
+
+    # Check if the folder of out_csv exists, if not create it
+    out_csv.parent.mkdir(parents=True, exist_ok=True)
+    df.to_csv(out_csv, index=False)
+
+
+if __name__ == "__main__":
+    main()