Precision Imaging for Intraindividual Investigation of the Reward Response.

Code ↔ Paper

7 matches between paragraphs of the paper and lines of its authors' code, computed by the harvester (lexical-v1). Click a colored paragraph or line to see its counterpart.

The 7 matches · 1 of them tie a paragraph to a whole file, not to given lines: a weak match, whose lines are not tinted

[1] § Methods › Neuroimaging Acquisition ↔ code/audit_neuromelanin.py, lines 746–875 · score 0.92 · magnetization transfer, partial Fourier, flip angle, gradient echo, echo sequence, coil
[2] § Methods › fMRI Tasks ↔ stimuli/sharedreward/genTrialList.m, the whole file · a weak match · score 0.70 · decision phase, 2550 ms, 850 ms, stranger, punishment, computer
[3] § Methods › fMRI Tasks ↔ stimuli/SharedReward_5button.py, lines 266–335 · score 0.68 · 1000–4000 ms, outcome phase, Shared Reward, card, money, feedback
[4] § Methods ↔ code/audit_neuromelanin.py, lines 1444–1555 · score 0.56 · OpenNeuro, Night Owls, NOSC, BIDS, Scan
[5] § Methods › fMRI Tasks ↔ stimuli/MID_5button.py, lines 520–587 · score 0.55 · Trial earnings, reward trials, cue, money, win, MID
[6] § Methods › fMRI Tasks ↔ stimuli/MID_practice.py, lines 485–548 · score 0.55 · Trial earnings, reward trials, cue, money, win, MID
[7] § Methods › fMRI Analysis and Moderating Factors ↔ masks/resample_to_study_mni_grid.sh, lines 1–77 · score 0.50 · NAcc, binarized, binary, thresholded, map, mask

Paper

Loaded from Europe PMC by your browser, not stored by OSCR: doi.org · Europe PMC

The paper is loaded when this pane is shown.

The paper is loaded by your browser, with JavaScript: read it at doi.org or on Europe PMC.

The authors' code

Python · 1,677 lines · 63 KB · MIT · 2 matches

#!/usr/bin/env python3
"""Audit raw Night Owls neuromelanin DICOM series without modifying source data.
The script reads one representative header from every raw DICOM series, identifies
neuromelanin candidates through both names and acquisition signatures, fingerprints
scientifically meaningful settings, and reconciles raw series against BIDS scans.tsv
inventories. Pixel data are never read. DICOM UIDs, dates, times, paths, and patient
fields are never written; dates and times are used only transiently for reconciliation.
"""
from __future__ import annotations
import argparse
import csv
import datetime as dt
import hashlib
import json
import logging
import math
import os
import re
import sys
from collections import Counter, defaultdict
from pathlib import Path
from typing import Any, Iterable, Mapping, Sequence
try:
import pydicom
except ImportError: # pragma: no cover - exercised by CLI setup failures
pydicom = None
LOG = logging.getLogger("neuromelanin_audit")
NA = "n/a"
# Deliberately excludes the generic token "NM", which has many false positives
# unless it is supported by acquisition parameters.
DEFAULT_NM_PATTERN = re.compile(
r"neuromelanin|neuro[\W_]*melanin|\bnmri\b|substantia[\W_]*nigra",
re.IGNORECASE,
)
PLAUSIBLE_PATTERN = re.compile(
r"\bt2[\W_]*star\b|t2\*|magnetization[\W_]*transfer|"
r"\bmt[\W_]*(?:gre|flash|tfl)\b",
re.IGNORECASE,
)
GENERIC_NM_PATTERN = re.compile(r"(?:^|[\W_])nm(?:$|[\W_])", re.IGNORECASE)
LIKELY_INTENTIONAL_EXCLUSION_PATTERN = re.compile(
r"localizer|scout|phoenix|survey|t1[\W_]*fl2d|setter",
re.IGNORECASE,
)
SOURCE_SESSION_PATTERN = re.compile(
r"Smith-NOSC-(?P<participant>[A-Za-z0-9]+)-SES(?P<session>[A-Za-z0-9]+)",
re.IGNORECASE,
)
BIDS_SESSION_PATTERN = re.compile(
r"sub-(?P<participant>[A-Za-z0-9]+).*?ses-(?P<session>[A-Za-z0-9]+)",
re.IGNORECASE,
)
MANIFEST_FIELDS = [
"participant_id",
"session_id",
"series_number",
"protocol_name",
"series_description",
"source_series_name",
"candidate_reason",
"sequence_version",
"sequence_fingerprint_sha256",
"n_dicoms",
"expected_even_session",
"audit_status",
"manufacturer",
"manufacturer_model",
"software_versions",
"magnetic_field_strength_T",
"sequence_name",
"pulse_sequence_name",
"echo_pulse_sequence",
"steady_state_pulse_sequence",
"scanning_sequence",
"sequence_variant",
"scan_options",
"mr_acquisition_type",
"repetition_time_ms",
"echo_time_ms",
"inversion_time_ms",
"flip_angle_deg",
"echo_train_length",
"rf_echo_train_length",
"gradient_echo_train_length",
"number_of_averages",
"slice_thickness_mm",
"spacing_between_slices_mm",
"pixel_spacing_row_mm",
"pixel_spacing_col_mm",
"rows",
"columns",
"acquisition_matrix",
"acquisition_frequency_encoding_steps",
"acquisition_phase_encoding_steps_in_plane",
"acquisition_phase_encoding_steps_out_of_plane",
"percent_sampling",
"percent_phase_fov",
"pixel_bandwidth_hz",
"receive_coil_name",
"in_plane_phase_encoding_direction",
"images_in_acquisition",
"number_of_frames",
"number_of_slices",
"estimated_slice_coverage_mm",
"number_of_volumes",
"number_of_multiframe_instances",
"parallel_reduction_factor_in_plane",
"parallel_reduction_factor_out_of_plane",
"parallel_acquisition_technique",
"partial_fourier",
"partial_fourier_direction",
"magnetization_transfer",
"siemens_mt_pulse_count",
"siemens_mt_frequency_offset_hz",
"siemens_mt_flip_angle_deg",
"siemens_mt_duration_us",
]
FINGERPRINT_FIELDS = [
"sequence_name",
"pulse_sequence_name",
"echo_pulse_sequence",
"steady_state_pulse_sequence",
"scanning_sequence",
"sequence_variant",
"scan_options",
"mr_acquisition_type",
"repetition_time_ms",
"echo_time_ms",
"inversion_time_ms",
"flip_angle_deg",
"echo_train_length",
"rf_echo_train_length",
"gradient_echo_train_length",
"number_of_averages",
"slice_thickness_mm",
"spacing_between_slices_mm",
"pixel_spacing_row_mm",
"pixel_spacing_col_mm",
"rows",
"columns",
"acquisition_matrix",
"acquisition_frequency_encoding_steps",
"acquisition_phase_encoding_steps_in_plane",
"acquisition_phase_encoding_steps_out_of_plane",
"percent_sampling",
"percent_phase_fov",
"pixel_bandwidth_hz",
"in_plane_phase_encoding_direction",
"number_of_slices",
"estimated_slice_coverage_mm",
"number_of_volumes",
"number_of_multiframe_instances",
"parallel_reduction_factor_in_plane",
"parallel_reduction_factor_out_of_plane",
"parallel_acquisition_technique",
"partial_fourier",
"partial_fourier_direction",
"magnetization_transfer",
"siemens_mt_pulse_count",
"siemens_mt_frequency_offset_hz",
"siemens_mt_flip_angle_deg",
"siemens_mt_duration_us",
]
SUMMARY_FIELDS = [
"sequence_version",
"sequence_fingerprint_sha256",
"n_series",
"n_sessions",
"n_participants",
"participants",
"sessions",
"protocol_names",
"series_descriptions",
] + [
field
for field in MANIFEST_FIELDS
if field
in {
"manufacturer",
"manufacturer_model",
"software_versions",
"magnetic_field_strength_T",
*FINGERPRINT_FIELDS,
"receive_coil_name",
"images_in_acquisition",
"number_of_frames",
}
]
INVENTORY_FIELDS = [
"participant_id",
"session_id",
"session_parity",
"nm_expected_from_design",
"nm_present",
"n_nm_series",
"sequence_version",
"status",
"notes",
]
REVIEW_FIELDS = [
"participant_id",
"session_id",
"series_number",
"protocol_name",
"series_description",
"source_series_name",
"review_reason",
"nm_signature_score",
"signature_matches",
"bids_counterpart_status",
"software_versions",
"sequence_name",
"pulse_sequence_name",
"scanning_sequence",
"mr_acquisition_type",
"repetition_time_ms",
"echo_time_ms",
"flip_angle_deg",
"slice_thickness_mm",
"spacing_between_slices_mm",
"pixel_spacing_row_mm",
"pixel_spacing_col_mm",
"rows",
"columns",
"number_of_slices",
"estimated_slice_coverage_mm",
"number_of_volumes",
"number_of_multiframe_instances",
]
RECONCILIATION_FIELDS = [
"participant_id",
"session_id",
"session_parity",
"series_number",
"source_series_name",
"protocol_name",
"series_description",
"n_dicoms",
"bids_counterpart_status",
"bids_match_evidence",
"likely_intentional_exclusion",
"nm_name_match",
"nm_signature_score",
"signature_matches",
"software_versions",
"sequence_name",
"pulse_sequence_name",
"scanning_sequence",
"sequence_variant",
"scan_options",
"mr_acquisition_type",
"repetition_time_ms",
"echo_time_ms",
"flip_angle_deg",
"slice_thickness_mm",
"spacing_between_slices_mm",
"pixel_spacing_row_mm",
"pixel_spacing_col_mm",
"rows",
"columns",
"number_of_slices",
"estimated_slice_coverage_mm",
"number_of_volumes",
"number_of_multiframe_instances",
]
OMISSION_SUMMARY_FIELDS = [
"software_versions",
"protocol_name",
"series_description",
"source_series_labels",
"sequence_name",
"pulse_sequence_name",
"repetition_time_ms",
"echo_time_ms",
"flip_angle_deg",
"slice_thickness_mm",
"spacing_between_slices_mm",
"rows",
"columns",
"number_of_slices",
"estimated_slice_coverage_mm",
"number_of_volumes",
"number_of_multiframe_instances",
"nm_signature_score",
"signature_matches",
"likely_intentional_exclusion",
"n_series",
"n_sessions",
"n_even_sessions",
"n_odd_sessions",
"participants",
"sessions",
]
PRIVATE_MT_KEYS = {
"siemens_mt_pulse_count": (
"sPrepPulses.lNoOfMTCPulses",
"sPrepPulses.lNoOfMtcPulses",
),
"siemens_mt_frequency_offset_hz": (
"sPrepPulses.dMTCFrequencyOffset",
"sPrepPulses.dMTFrequencyOffset",
),
"siemens_mt_flip_angle_deg": (
"sPrepPulses.dMTFlipAngle",
"sPrepPulses.dMTCFlipAngle",
),
"siemens_mt_duration_us": (
"sPrepPulses.dMTDuration",
"sPrepPulses.dMTCDuration",
),
}
def natural_key(value: str) -> tuple[Any, ...]:
"""Return a deterministic human-friendly key for identifiers and labels."""
return tuple(int(part) if part.isdigit() else part.lower() for part in re.split(r"(\d+)", value))
def clean_text(value: Any) -> str:
"""Convert a DICOM value to one safe, single-line TSV cell."""
if value is None or value == "":
return NA
if isinstance(value, bytes):
value = value.decode("latin-1", errors="replace")
if isinstance(value, (list, tuple)) or value.__class__.__name__ == "MultiValue":
value = "\\".join(str(item) for item in value)
text = str(value).replace("\t", " ").replace("\r", " ").replace("\n", " ")
text = re.sub(r"\s+", " ", text).strip()
return text if text else NA
def normalized_number(value: Any, digits: int = 6) -> str:
"""Normalize numerical DICOM values so trivial float formatting does not split versions."""
if value is None or value == "":
return NA
try:
number = float(value)
except (TypeError, ValueError):
return clean_text(value)
if not math.isfinite(number):
return NA
if abs(number - round(number)) < 10 ** (-digits):
return str(int(round(number)))
return f"{number:.{digits}f}".rstrip("0").rstrip(".")
def normalize_identifier(prefix: str, value: str) -> str:
value = re.sub(rf"^{re.escape(prefix)}-", "", value, flags=re.IGNORECASE)
if value.isdigit():
value = value.zfill(2 if prefix == "ses" else 3)
return f"{prefix}-{value}"
def parse_session_from_path(path: Path) -> tuple[str, str] | None:
joined = "/".join(path.parts)
match = SOURCE_SESSION_PATTERN.search(joined) or BIDS_SESSION_PATTERN.search(joined)
if not match:
return None
return (
normalize_identifier("sub", match.group("participant")),
normalize_identifier("ses", match.group("session")),
)
def session_number(session_id: str) -> int | None:
match = re.search(r"(\d+)$", session_id)
return int(match.group(1)) if match else None
def expected_even(session_id: str) -> bool | None:
number = session_number(session_id)
return None if number is None else number % 2 == 0
def yes_no(value: bool | None) -> str:
if value is None:
return NA
return "yes" if value else "no"
def number(value: str) -> float | None:
if value == NA:
return None
try:
result = float(value)
except (TypeError, ValueError):
return None
return result if math.isfinite(result) else None
def near(value: str, target: float, tolerance: float) -> bool:
parsed = number(value)
return parsed is not None and abs(parsed - target) <= tolerance
def acquisition_clock_seconds(dataset: Any) -> float | None:
"""Return month/day/time as a year-independent scalar for private in-memory matching."""
date_value = get_value(dataset, "AcquisitionDate") or get_value(dataset, "SeriesDate")
time_value = get_value(dataset, "AcquisitionTime") or get_value(dataset, "SeriesTime")
combined = clean_text(get_value(dataset, "AcquisitionDateTime"))
if combined != NA and len(re.sub(r"\D", "", combined.split(".", 1)[0])) >= 14:
date_value = combined[:8]
time_value = combined[8:]
if not date_value or not time_value:
return None
date_digits = re.sub(r"\D", "", str(date_value))
time_text = str(time_value).strip()
match = re.match(r"(?P<h>\d{2})(?P<m>\d{2})(?P<s>\d{2}(?:\.\d+)?)", time_text)
if len(date_digits) < 8 or not match:
return None
try:
anchor = dt.datetime(2000, int(date_digits[4:6]), int(date_digits[6:8]))
seconds = (
int(match.group("h")) * 3600
+ int(match.group("m")) * 60
+ float(match.group("s"))
)
except (TypeError, ValueError):
return None
return (anchor - dt.datetime(2000, 1, 1)).days * 86400 + seconds
def bids_clock_seconds(value: str) -> float | None:
try:
parsed = dt.datetime.fromisoformat(value.strip().replace("Z", "+00:00"))
anchor = dt.datetime(2000, parsed.month, parsed.day)
except (TypeError, ValueError):
return None
return (
(anchor - dt.datetime(2000, 1, 1)).days * 86400
+ parsed.hour * 3600
+ parsed.minute * 60
+ parsed.second
+ parsed.microsecond / 1_000_000
)
def neuromelanin_signature(row: Mapping[str, str]) -> tuple[int, str]:
"""Score name-independent similarity to the observed E11/XA30 NM acquisition."""
matches: list[str] = []
score = 0
if near(row.get("repetition_time_ms", NA), 641, 25):
score += 2
matches.append("TR_about_641ms")
if near(row.get("echo_time_ms", NA), 3.97, 0.35):
score += 2
matches.append("TE_about_3.97ms")
if near(row.get("flip_angle_deg", NA), 50, 5):
score += 2
matches.append("flip_angle_about_50deg")
row_spacing = number(row.get("pixel_spacing_row_mm", NA))
col_spacing = number(row.get("pixel_spacing_col_mm", NA))
if (
row_spacing is not None
and col_spacing is not None
and 0.55 <= row_spacing <= 0.75
and 0.55 <= col_spacing <= 0.75
):
score += 1
matches.append("in_plane_resolution_about_0.64mm")
sequence_text = " ".join(
row.get(field, NA)
for field in (
"sequence_name",
"pulse_sequence_name",
"scanning_sequence",
"mr_acquisition_type",
)
).lower()
if "fl2d1" in sequence_text or (
"gradient" in sequence_text and "2d" in sequence_text
):
score += 1
matches.append("2D_gradient_echo")
if any(row.get(field, NA) == "6" for field in (
"number_of_averages",
"number_of_volumes",
"number_of_multiframe_instances",
)):
score += 1
matches.append("six_repeats")
if row.get("number_of_slices", NA) == "24":
score += 1
matches.append("24_slices")
return score, "+".join(matches) if matches else NA
def estimated_slice_coverage(row: Mapping[str, str]) -> str:
"""Estimate first-to-last outer-edge coverage from slice count and spacing."""
slices = number(row.get("number_of_slices", NA))
thickness = number(row.get("slice_thickness_mm", NA))
spacing = number(row.get("spacing_between_slices_mm", NA))
if slices is None or slices < 1 or thickness is None:
return NA
step = spacing if spacing is not None else thickness
return normalized_number(thickness + max(0, slices - 1) * step)
def source_series_label(path: Path) -> str:
"""Return a scan-folder label for matching only; it is never emitted to output."""
parts = list(path.parts)
lowered = [part.lower() for part in parts]
if "scans" in lowered:
index = len(lowered) - 1 - lowered[::-1].index("scans")
if index + 1 < len(parts):
return parts[index + 1]
return path.parent.parent.parent.name if len(parts) >= 4 else path.name
def discover_series(source_root: Path) -> tuple[list[tuple[Path, list[Path]]], set[tuple[str, str]]]:
"""Discover historical XNAT DICOM resource directories and all source sessions."""
series: list[tuple[Path, list[Path]]] = []
sessions: set[tuple[str, str]] = set()
for current, dirs, files in os.walk(source_root):
current_path = Path(current)
parsed = parse_session_from_path(current_path)
if parsed:
sessions.add(parsed)
if not files:
continue
parent_names = [part.lower() for part in current_path.parts[-3:]]
is_dicom_resource = (
current_path.name.lower() == "files"
and "dicom" in parent_names
) or current_path.name.lower() == "dicom"
if not is_dicom_resource:
continue
paths = sorted(
(current_path / name for name in files if (current_path / name).is_file()),
key=lambda item: natural_key(item.name),
)
if paths:
series.append((current_path, paths))
dirs[:] = []
series.sort(key=lambda item: natural_key(str(item[0])))
return series, sessions
def read_representative(files: Sequence[Path]) -> tuple[Any | None, int]:
"""Read the first valid DICOM header and return it with the unreadable-file count."""
unreadable = 0
for path in files:
try:
dataset = pydicom.dcmread(path, stop_before_pixels=True, force=True)
except Exception:
unreadable += 1
continue
if any(
getattr(dataset, keyword, None)
for keyword in ("Modality", "SOPClassUID", "ProtocolName", "SeriesDescription")
):
return dataset, unreadable
unreadable += 1
return None, unreadable
def get_value(dataset: Any, keyword: str) -> Any:
value = getattr(dataset, keyword, None)
return value.value if hasattr(value, "value") else value
def get_tag_value(dataset: Any, group: int, element: int) -> Any:
entry = dataset.get((group, element))
return entry.value if entry is not None else None
def functional_group_roots(dataset: Any) -> Iterable[Any]:
"""Yield shared and one representative per-frame functional-group item."""
shared = get_value(dataset, "SharedFunctionalGroupsSequence")
if shared:
yield shared[0]
per_frame = get_value(dataset, "PerFrameFunctionalGroupsSequence")
if per_frame:
yield per_frame[0]
def acquisition_value(
dataset: Any,
keyword: str,
sequence_keywords: Sequence[str] = (),
) -> Any:
"""Read a top-level value or its Enhanced MR functional-group equivalent."""
value = get_value(dataset, keyword)
if value not in (None, ""):
return value
for root in functional_group_roots(dataset):
for sequence_keyword in sequence_keywords:
sequence = get_value(root, sequence_keyword)
if not sequence:
continue
value = get_value(sequence[0], keyword)
if value not in (None, ""):
return value
return None
def private_text(dataset: Any) -> str:
"""Read only Siemens CSA blobs in memory; never return or write the full contents."""
chunks: list[str] = []
for tag in ((0x0029, 0x1010), (0x0029, 0x1020)):
value = get_tag_value(dataset, *tag)
if isinstance(value, bytes):
chunks.append(value.decode("latin-1", errors="ignore"))
elif value:
chunks.append(str(value))
return "\n".join(chunks)
def extract_private_mt(dataset: Any) -> dict[str, str]:
text = private_text(dataset)
values = {field: NA for field in PRIVATE_MT_KEYS}
if not text:
return values
for field, keys in PRIVATE_MT_KEYS.items():
for key in keys:
match = re.search(
rf"(?m){re.escape(key)}\s*=\s*([^\x00\r\n]+)",
text,
)
if match:
raw = match.group(1).split("#", 1)[0].strip()
values[field] = normalized_number(raw)
break
return values
def acquisition_matrix(dataset: Any) -> str:
value = acquisition_value(dataset, "AcquisitionMatrix", ("MRFOVGeometrySequence",))
if value is None:
return NA
try:
return "x".join(str(int(item)) for item in value)
except (TypeError, ValueError):
return clean_text(value)
def pixel_spacing(dataset: Any) -> tuple[str, str]:
value = acquisition_value(dataset, "PixelSpacing", ("PixelMeasuresSequence",))
try:
return normalized_number(value[0]), normalized_number(value[1])
except (TypeError, IndexError):
return NA, NA
def infer_number_of_slices(dataset: Any, n_dicoms: int) -> str:
images = get_value(dataset, "ImagesInAcquisition")
frames = get_value(dataset, "NumberOfFrames")
if images not in (None, ""):
return normalized_number(images)
if frames not in (None, "") and normalized_number(frames) != "1":
return normalized_number(frames)
return NA
def infer_multiframe_instances(dataset: Any, n_dicoms: int) -> str:
frames = normalized_number(get_value(dataset, "NumberOfFrames"))
if frames not in (NA, "1"):
return str(n_dicoms)
return NA
def infer_series_layout(files: Sequence[Path], dataset: Any) -> dict[str, str]:
"""Infer slice and volume counts without treating all 2D instances as slices."""
frames = normalized_number(get_value(dataset, "NumberOfFrames"))
if frames not in (NA, "1"):
return {
"number_of_slices": frames,
"number_of_volumes": str(len(files)),
"number_of_multiframe_instances": str(len(files)),
}
positions: set[tuple[float, ...]] = set()
slice_locations: set[float] = set()
temporal_positions: set[str] = set()
tags = [
"ImagePositionPatient",
"SliceLocation",
"TemporalPositionIdentifier",
"AcquisitionNumber",
]
for path in files:
try:
header = pydicom.dcmread(
path,
stop_before_pixels=True,
force=True,
specific_tags=tags,
)
except Exception:
continue
position = get_value(header, "ImagePositionPatient")
if position:
try:
positions.add(tuple(round(float(value), 4) for value in position))
except (TypeError, ValueError):
pass
location = get_value(header, "SliceLocation")
if location not in (None, ""):
try:
slice_locations.add(round(float(location), 4))
except (TypeError, ValueError):
pass
temporal = get_value(header, "TemporalPositionIdentifier")
if temporal in (None, ""):
temporal = get_value(header, "AcquisitionNumber")
if temporal not in (None, ""):
temporal_positions.add(clean_text(temporal))
n_slices = len(positions) or len(slice_locations)
if not n_slices:
return {
"number_of_slices": infer_number_of_slices(dataset, len(files)),
"number_of_volumes": (
str(len(temporal_positions)) if temporal_positions else NA
),
"number_of_multiframe_instances": NA,
}
if temporal_positions:
n_volumes = len(temporal_positions)
elif len(files) % n_slices == 0:
n_volumes = len(files) // n_slices
else:
n_volumes = 0
return {
"number_of_slices": str(n_slices),
"number_of_volumes": str(n_volumes) if n_volumes else NA,
"number_of_multiframe_instances": NA,
}
def extract_metadata(dataset: Any, n_dicoms: int) -> dict[str, str]:
row_spacing, col_spacing = pixel_spacing(dataset)
timing_sequences = ("MRTimingAndRelatedParametersSequence",)
echo_sequences = ("MREchoSequence",)
modifier_sequences = ("MRModifierSequence", "MRImagingModifierSequence")
fov_sequences = ("MRFOVGeometrySequence",)
imaging_modifier_sequences = (
"MRImagingModifierSequence",
"MRModifierSequence",
"MRFOVGeometrySequence",
)
metadata = {
"series_number": normalized_number(get_value(dataset, "SeriesNumber")),
"protocol_name": clean_text(get_value(dataset, "ProtocolName")),
"series_description": clean_text(get_value(dataset, "SeriesDescription")),
"manufacturer": clean_text(get_value(dataset, "Manufacturer")),
"manufacturer_model": clean_text(get_value(dataset, "ManufacturerModelName")),
"software_versions": clean_text(get_value(dataset, "SoftwareVersions")),
"magnetic_field_strength_T": normalized_number(get_value(dataset, "MagneticFieldStrength")),
"sequence_name": clean_text(get_value(dataset, "SequenceName")),
"pulse_sequence_name": clean_text(
acquisition_value(dataset, "PulseSequenceName", modifier_sequences)
),
"echo_pulse_sequence": clean_text(
acquisition_value(dataset, "EchoPulseSequence", modifier_sequences)
),
"steady_state_pulse_sequence": clean_text(
acquisition_value(dataset, "SteadyStatePulseSequence", modifier_sequences)
),
"scanning_sequence": clean_text(get_value(dataset, "ScanningSequence")),
"sequence_variant": clean_text(get_value(dataset, "SequenceVariant")),
"scan_options": clean_text(get_value(dataset, "ScanOptions")),
"mr_acquisition_type": clean_text(get_value(dataset, "MRAcquisitionType")),
"repetition_time_ms": normalized_number(
acquisition_value(dataset, "RepetitionTime", timing_sequences)
),
"echo_time_ms": normalized_number(
acquisition_value(dataset, "EchoTime", echo_sequences)
or acquisition_value(dataset, "EffectiveEchoTime", echo_sequences)
),
"inversion_time_ms": normalized_number(
acquisition_value(dataset, "InversionTime", modifier_sequences)
or acquisition_value(dataset, "InversionTimes", modifier_sequences)
),
"flip_angle_deg": normalized_number(
acquisition_value(dataset, "FlipAngle", timing_sequences)
),
"echo_train_length": normalized_number(
acquisition_value(dataset, "EchoTrainLength", timing_sequences)
),
"rf_echo_train_length": normalized_number(
acquisition_value(dataset, "RFEchoTrainLength", timing_sequences)
),
"gradient_echo_train_length": normalized_number(
acquisition_value(dataset, "GradientEchoTrainLength", timing_sequences)
),
"number_of_averages": normalized_number(
acquisition_value(dataset, "NumberOfAverages", ("MRAveragesSequence",))
),
"slice_thickness_mm": normalized_number(
acquisition_value(dataset, "SliceThickness", ("PixelMeasuresSequence",))
),
"spacing_between_slices_mm": normalized_number(
acquisition_value(dataset, "SpacingBetweenSlices", ("PixelMeasuresSequence",))
),
"pixel_spacing_row_mm": row_spacing,
"pixel_spacing_col_mm": col_spacing,
"rows": normalized_number(get_value(dataset, "Rows")),
"columns": normalized_number(get_value(dataset, "Columns")),
"acquisition_matrix": acquisition_matrix(dataset),
"acquisition_frequency_encoding_steps": normalized_number(
acquisition_value(dataset, "MRAcquisitionFrequencyEncodingSteps", fov_sequences)
),
"acquisition_phase_encoding_steps_in_plane": normalized_number(
acquisition_value(dataset, "MRAcquisitionPhaseEncodingStepsInPlane", fov_sequences)
),
"acquisition_phase_encoding_steps_out_of_plane": normalized_number(
acquisition_value(dataset, "MRAcquisitionPhaseEncodingStepsOutOfPlane", fov_sequences)
),
"percent_sampling": normalized_number(
acquisition_value(dataset, "PercentSampling", fov_sequences)
),
"percent_phase_fov": normalized_number(
acquisition_value(dataset, "PercentPhaseFieldOfView", fov_sequences)
),
"pixel_bandwidth_hz": normalized_number(get_value(dataset, "PixelBandwidth")),
"receive_coil_name": clean_text(
acquisition_value(dataset, "ReceiveCoilName", ("MRReceiveCoilSequence",))
),
"in_plane_phase_encoding_direction": clean_text(
acquisition_value(dataset, "InPlanePhaseEncodingDirection", fov_sequences)
),
"images_in_acquisition": normalized_number(get_value(dataset, "ImagesInAcquisition")),
"number_of_frames": normalized_number(get_value(dataset, "NumberOfFrames")),
"number_of_slices": infer_number_of_slices(dataset, n_dicoms),
"number_of_volumes": NA,
"number_of_multiframe_instances": infer_multiframe_instances(dataset, n_dicoms),
"parallel_reduction_factor_in_plane": normalized_number(
acquisition_value(
dataset,
"ParallelReductionFactorInPlane",
imaging_modifier_sequences,
)
),
"parallel_reduction_factor_out_of_plane": normalized_number(
acquisition_value(
dataset,
"ParallelReductionFactorOutOfPlane",
imaging_modifier_sequences,
)
),
"parallel_acquisition_technique": clean_text(
acquisition_value(
dataset,
"ParallelAcquisitionTechnique",
imaging_modifier_sequences,
)
),
"partial_fourier": normalized_number(
acquisition_value(dataset, "PartialFourier", imaging_modifier_sequences)
),
"partial_fourier_direction": clean_text(
acquisition_value(dataset, "PartialFourierDirection", imaging_modifier_sequences)
),
"magnetization_transfer": clean_text(
acquisition_value(dataset, "MagnetizationTransfer", modifier_sequences)
),
}
metadata.update(extract_private_mt(dataset))
return metadata
def candidate_reason(protocol: str, description: str, source_label: str) -> str | None:
sources = []
if DEFAULT_NM_PATTERN.search(protocol):
sources.append("protocol_name")
if DEFAULT_NM_PATTERN.search(description):
sources.append("series_description")
if DEFAULT_NM_PATTERN.search(source_label):
sources.append("source_series_name")
return "+".join(sources) if sources else None
def fingerprint(row: Mapping[str, str]) -> tuple[str, str]:
content = {field: row.get(field, NA) for field in FINGERPRINT_FIELDS}
serialized = json.dumps(content, sort_keys=True, separators=(",", ":"))
return serialized, hashlib.sha256(serialized.encode("utf-8")).hexdigest()
def discover_bids_sessions(bids_root: Path) -> set[tuple[str, str]]:
sessions: set[tuple[str, str]] = set()
if not bids_root.exists():
LOG.warning("BIDS root does not exist; source-only inventory will be generated")
return sessions
for participant_dir in bids_root.glob("sub-*"):
if not participant_dir.is_dir():
continue
for session_dir in participant_dir.glob("ses-*"):
if session_dir.is_dir():
sessions.add((participant_dir.name, session_dir.name))
return sessions
def audit_series(
source_root: Path,
) -> tuple[
list[dict[str, str]],
set[tuple[str, str]],
int,
int,
]:
discovered, source_sessions = discover_series(source_root)
if not discovered:
raise RuntimeError("No raw DICOM series directories were found under source root")
rows: list[dict[str, str]] = []
unreadable_series = 0
unparsed_series = 0
for index, (directory, files) in enumerate(discovered, start=1):
if index % 100 == 0:
LOG.info("Inspected %d/%d series", index, len(discovered))
parsed = parse_session_from_path(directory)
dataset, unreadable_files = read_representative(files)
label = source_series_label(directory)
if dataset is None:
unreadable_series += 1
participant_id, session_id = parsed if parsed else (NA, NA)
row = {field: NA for field in MANIFEST_FIELDS}
row.update(
{
"participant_id": participant_id,
"session_id": session_id,
"candidate_reason": (
"source_series_name" if DEFAULT_NM_PATTERN.search(label) else NA
),
"source_series_name": label,
"n_dicoms": str(len(files)),
"expected_even_session": (
yes_no(expected_even(session_id)) if parsed else NA
),
"audit_status": "header_unreadable",
"_series_instance_uid": NA,
"_acquisition_clock_seconds": None,
}
)
rows.append(row)
continue
metadata = extract_metadata(dataset, len(files))
metadata.update(infer_series_layout(files, dataset))
metadata["estimated_slice_coverage_mm"] = estimated_slice_coverage(metadata)
protocol = metadata["protocol_name"]
description = metadata["series_description"]
if not parsed:
unparsed_series += 1
participant_id, session_id = parsed if parsed else (NA, NA)
reason = candidate_reason(protocol, description, label)
row = {field: NA for field in MANIFEST_FIELDS}
row.update(metadata)
row.update(
{
"participant_id": participant_id,
"session_id": session_id,
"candidate_reason": reason or NA,
"source_series_name": label,
"n_dicoms": str(len(files)),
"expected_even_session": (
yes_no(expected_even(session_id)) if parsed else NA
),
"audit_status": (
"ok" if unreadable_files == 0 else "representative_read_after_error"
),
"_series_instance_uid": clean_text(
get_value(dataset, "SeriesInstanceUID")
),
"_acquisition_clock_seconds": acquisition_clock_seconds(dataset),
}
)
score, matches = neuromelanin_signature(row)
row["nm_signature_score"] = str(score)
row["signature_matches"] = matches
rows.append(row)
rows.sort(
key=lambda row: (
natural_key(row["participant_id"]),
natural_key(row["session_id"]),
natural_key(row["series_number"]),
row["protocol_name"].lower(),
)
)
LOG.info(
"Inspected %d series without a parseable participant/session",
unparsed_series,
)
return rows, source_sessions, unreadable_series, unparsed_series
def deduplicate_resource_copies(
rows: Sequence[dict[str, str]],
) -> tuple[list[dict[str, str]], int, int]:
"""Collapse archive-resource copies sharing a Series Instance UID in one session.
UIDs are used only in memory and are removed before any output is written. When
copies differ in file count, the most complete resource is retained and flagged.
"""
grouped: dict[tuple[str, str, str], list[dict[str, str]]] = defaultdict(list)
ungrouped: list[dict[str, str]] = []
for row in rows:
uid = row.get("_series_instance_uid", NA)
if uid == NA:
ungrouped.append(row)
else:
grouped[(row["participant_id"], row["session_id"], uid)].append(row)
retained = list(ungrouped)
duplicate_resources = 0
incomplete_copy_groups = 0
for group in grouped.values():
counts = {int(row["n_dicoms"]) for row in group}
ranked = sorted(
group,
key=lambda row: (
-int(row["n_dicoms"]),
-sum(row.get(field, NA) != NA for field in FINGERPRINT_FIELDS),
json.dumps(
{field: row.get(field, NA) for field in MANIFEST_FIELDS},
sort_keys=True,
),
),
)
selected = ranked[0]
if len(group) > 1:
duplicate_resources += len(group) - 1
statuses = [] if selected["audit_status"] == "ok" else selected["audit_status"].split(";")
statuses.append(f"duplicate_source_resources_excluded={len(group) - 1}")
if len(counts) > 1:
incomplete_copy_groups += 1
statuses.append("incomplete_source_copy_excluded")
selected["audit_status"] = ";".join(dict.fromkeys(statuses))
retained.append(selected)
for row in retained:
row.pop("_series_instance_uid", None)
retained.sort(
key=lambda row: (
natural_key(row["participant_id"]),
natural_key(row["session_id"]),
natural_key(row["series_number"]),
row["protocol_name"].lower(),
)
)
return retained, duplicate_resources, incomplete_copy_groups
def discover_bids_scan_index(
bids_root: Path,
) -> dict[tuple[str, str], list[tuple[float, str]]]:
"""Index deidentified BIDS scan timestamps without emitting dates or times."""
index: dict[tuple[str, str], list[tuple[float, str]]] = defaultdict(list)
if not bids_root.exists():
return index
for path in sorted(bids_root.glob("sub-*/ses-*/*_scans.tsv")):
parsed_session = parse_session_from_path(path)
if not parsed_session:
continue
try:
with path.open(encoding="utf-8", newline="") as stream:
for item in csv.DictReader(stream, delimiter="\t"):
clock = bids_clock_seconds(item.get("acq_time", ""))
filename = item.get("filename", "")
if clock is not None:
index[parsed_session].append((clock, filename))
except OSError:
LOG.warning("Could not read one BIDS scans.tsv inventory")
return index
def reconcile_source_to_bids(
rows: Sequence[dict[str, str]],
bids_sessions: set[tuple[str, str]],
bids_index: Mapping[tuple[str, str], Sequence[tuple[float, str]]],
tolerance_seconds: float = 2.0,
) -> None:
"""Annotate raw series using time only in memory; never write the time itself."""
for row in rows:
participant_id = row["participant_id"]
session_id = row["session_id"]
session = (participant_id, session_id)
source_text = " ".join(
row.get(field, NA)
for field in ("protocol_name", "series_description", "source_series_name")
)
row["session_parity"] = (
"even" if expected_even(session_id) is True
else "odd" if expected_even(session_id) is False
else NA
)
row["likely_intentional_exclusion"] = yes_no(
bool(LIKELY_INTENTIONAL_EXCLUSION_PATTERN.search(source_text))
)
row["nm_name_match"] = yes_no(row.get("candidate_reason", NA) != NA)
if participant_id == NA or session_id == NA:
row["bids_counterpart_status"] = "session_unparseable"
row["bids_match_evidence"] = NA
continue
if session not in bids_sessions:
row["bids_counterpart_status"] = "bids_session_absent"
row["bids_match_evidence"] = NA
continue
candidates = bids_index.get(session, ())
if not candidates:
row["bids_counterpart_status"] = "bids_scan_inventory_unavailable"
row["bids_match_evidence"] = NA
continue
clock = row.get("_acquisition_clock_seconds")
if clock is None:
row["bids_counterpart_status"] = "match_indeterminate_no_source_time"
row["bids_match_evidence"] = NA
continue
matched = [
filename for candidate_clock, filename in candidates
if abs(candidate_clock - float(clock)) <= tolerance_seconds
]
if matched:
datatypes = sorted(
{filename.split("/", 1)[0] for filename in matched if "/" in filename},
key=natural_key,
)
row["bids_counterpart_status"] = "converted_or_listed"
row["bids_match_evidence"] = (
f"{len(matched)}_BIDS_file(s);datatype={','.join(datatypes) or NA}"
)
else:
row["bids_counterpart_status"] = "no_bids_counterpart"
row["bids_match_evidence"] = "no_scan_inventory_time_within_2s"
def build_candidate_sets(
rows: Sequence[dict[str, str]],
) -> tuple[list[dict[str, str]], list[dict[str, str]]]:
confirmed: list[dict[str, str]] = []
review: list[dict[str, str]] = []
for source_row in rows:
row = dict(source_row)
name_reason = row.get("candidate_reason", NA)
score = int(row.get("nm_signature_score", "0"))
source_text = " ".join(
row.get(field, NA)
for field in ("protocol_name", "series_description", "source_series_name")
)
if name_reason != NA and row["participant_id"] != NA and row["session_id"] != NA:
confirmed.append(row)
continue
reasons: list[str] = []
if name_reason != NA:
reasons.append("neuromelanin_candidate_without_parseable_session")
if score >= 6:
reasons.append("acquisition_signature_score_ge_6")
if PLAUSIBLE_PATTERN.search(source_text):
reasons.append("plausible_name_without_neuromelanin_label")
if GENERIC_NM_PATTERN.search(source_text) and score >= 2:
reasons.append("generic_NM_label_supported_by_signature")
if not reasons:
continue
review_row = {field: row.get(field, NA) for field in REVIEW_FIELDS}
review_row["review_reason"] = ";".join(reasons)
review.append(review_row)
confirmed.sort(
key=lambda row: (
natural_key(row["participant_id"]),
natural_key(row["session_id"]),
natural_key(row["series_number"]),
)
)
review.sort(
key=lambda row: (
natural_key(row["participant_id"]),
natural_key(row["session_id"]),
natural_key(row["series_number"]),
)
)
return confirmed, review
def build_omission_summary(
rows: Sequence[dict[str, str]],
) -> list[dict[str, str]]:
descriptive_fields = [
field
for field in OMISSION_SUMMARY_FIELDS
if field
not in {
"source_series_labels",
"n_series",
"n_sessions",
"n_even_sessions",
"n_odd_sessions",
"participants",
"sessions",
}
]
grouped: dict[tuple[str, ...], list[dict[str, str]]] = defaultdict(list)
for row in rows:
if row.get("bids_counterpart_status") != "no_bids_counterpart":
continue
key = tuple(row.get(field, NA) for field in descriptive_fields)
grouped[key].append(row)
summary: list[dict[str, str]] = []
for key, group in grouped.items():
item = dict(zip(descriptive_fields, key))
sessions = sorted(
{
f"{row['participant_id']}/{row['session_id']}"
for row in group
if row["participant_id"] != NA and row["session_id"] != NA
},
key=natural_key,
)
participants = sorted(
{row["participant_id"] for row in group if row["participant_id"] != NA},
key=natural_key,
)
item.update(
{
"source_series_labels": aggregate_values(group, "source_series_name"),
"n_series": str(len(group)),
"n_sessions": str(len(sessions)),
"n_even_sessions": str(sum(row["session_parity"] == "even" for row in group)),
"n_odd_sessions": str(sum(row["session_parity"] == "odd" for row in group)),
"participants": ",".join(participants) if participants else NA,
"sessions": ",".join(sessions) if sessions else NA,
}
)
summary.append(item)
summary.sort(
key=lambda row: (
-int(row["n_sessions"]),
-int(row["nm_signature_score"]),
natural_key(row["software_versions"]),
natural_key(row["protocol_name"]),
)
)
return summary
def assign_versions(rows: list[dict[str, str]]) -> None:
serialized_to_hash: dict[str, str] = {}
for row in rows:
if row["audit_status"] == "header_unreadable":
continue
serialized, digest = fingerprint(row)
serialized_to_hash[serialized] = digest
row["sequence_fingerprint_sha256"] = digest
versions = {
serialized: f"NM_v{index:02d}"
for index, serialized in enumerate(sorted(serialized_to_hash), start=1)
}
for row in rows:
if row["audit_status"] == "header_unreadable":
continue
serialized, _ = fingerprint(row)
row["sequence_version"] = versions[serialized]
counts = Counter((row["participant_id"], row["session_id"]) for row in rows)
for row in rows:
statuses = [] if row["audit_status"] == "ok" else row["audit_status"].split(";")
if expected_even(row["session_id"]) is False:
statuses.append("unexpected_odd_session")
if counts[(row["participant_id"], row["session_id"])] > 1:
statuses.append("multiple_nm_series")
row["audit_status"] = ";".join(dict.fromkeys(statuses)) if statuses else "ok"
def aggregate_values(rows: Sequence[Mapping[str, str]], field: str) -> str:
values = sorted({row.get(field, NA) for row in rows}, key=natural_key)
return " | ".join(values)
def build_summary(rows: Sequence[dict[str, str]]) -> list[dict[str, str]]:
grouped: dict[str, list[dict[str, str]]] = defaultdict(list)
for row in rows:
if row["sequence_version"] != NA:
grouped[row["sequence_version"]].append(row)
summary: list[dict[str, str]] = []
for version in sorted(grouped, key=natural_key):
group = grouped[version]
participants = sorted({row["participant_id"] for row in group}, key=natural_key)
sessions = sorted(
{f"{row['participant_id']}/{row['session_id']}" for row in group},
key=natural_key,
)
item = {field: NA for field in SUMMARY_FIELDS}
item.update(
{
"sequence_version": version,
"sequence_fingerprint_sha256": group[0]["sequence_fingerprint_sha256"],
"n_series": str(len(group)),
"n_sessions": str(len(sessions)),
"n_participants": str(len(participants)),
"participants": ",".join(participants),
"sessions": ",".join(sessions),
"protocol_names": aggregate_values(group, "protocol_name"),
"series_descriptions": aggregate_values(group, "series_description"),
}
)
for field in SUMMARY_FIELDS[9:]:
item[field] = aggregate_values(group, field)
summary.append(item)
return summary
def build_inventory(
rows: Sequence[dict[str, str]],
source_sessions: set[tuple[str, str]],
bids_sessions: set[tuple[str, str]],
) -> list[dict[str, str]]:
by_session: dict[tuple[str, str], list[dict[str, str]]] = defaultdict(list)
for row in rows:
by_session[(row["participant_id"], row["session_id"])].append(row)
inventory = []
for participant_id, session_id in sorted(
source_sessions | bids_sessions,
key=lambda item: (natural_key(item[0]), natural_key(item[1])),
):
scans = by_session[(participant_id, session_id)]
versions = sorted(
{row["sequence_version"] for row in scans if row["sequence_version"] != NA},
key=natural_key,
)
expected = expected_even(session_id)
in_source = (participant_id, session_id) in source_sessions
in_bids = (participant_id, session_id) in bids_sessions
notes = []
if in_source and not in_bids:
status = "bids_session_missing"
notes.append("source session present; corresponding BIDS session absent")
elif in_bids and not in_source:
status = "source_missing"
notes.append("BIDS session present; corresponding source session absent")
elif len(scans) > 1:
status = "multiple_nm_series"
notes.append("multiple neuromelanin series detected")
if expected is False:
notes.append("neuromelanin not expected in odd session")
elif scans and expected is True:
status = "expected_found"
elif scans and expected is False:
status = "unexpected_found"
elif not scans and expected is True:
status = "expected_missing"
elif not scans and expected is False:
status = "not_expected_not_found"
else:
status = "other"
notes.append("session parity is not numeric")
inventory.append(
{
"participant_id": participant_id,
"session_id": session_id,
"session_parity": (
"even" if expected is True else "odd" if expected is False else NA
),
"nm_expected_from_design": yes_no(expected),
"nm_present": yes_no(bool(scans)),
"n_nm_series": str(len(scans)),
"sequence_version": ",".join(versions) if versions else NA,
"status": status,
"notes": "; ".join(notes) if notes else NA,
}
)
return inventory
def write_tsv(path: Path, rows: Sequence[Mapping[str, str]], fields: Sequence[str]) -> None:
path.parent.mkdir(parents=True, exist_ok=True)
with path.open("w", encoding="utf-8", newline="") as stream:
writer = csv.DictWriter(stream, fieldnames=fields, delimiter="\t", lineterminator="\n")
writer.writeheader()
writer.writerows({field: row.get(field, NA) for field in fields} for row in rows)
def differing_fields(summary: Sequence[Mapping[str, str]]) -> list[str]:
descriptive = []
for field in FINGERPRINT_FIELDS:
values = {row.get(field, NA) for row in summary}
if len(values) > 1:
descriptive.append(field)
return descriptive
def markdown_table(summary: Sequence[Mapping[str, str]], fields: Sequence[str]) -> str:
columns = ["sequence_version", "n_series", "n_sessions", *fields]
header = "| " + " | ".join(columns) + " |"
rule = "| " + " | ".join("---" for _ in columns) + " |"
body = []
for row in summary:
cells = [str(row.get(column, NA)).replace("|", "\\|") for column in columns]
body.append("| " + " | ".join(cells) + " |")
return "\n".join((header, rule, *body))
def status_sessions(inventory: Sequence[Mapping[str, str]], status: str) -> str:
values = [
f"{row['participant_id']}/{row['session_id']}"
for row in inventory
if row["status"] == status
]
return ", ".join(values) if values else "None."
def label_parameter_relationships(rows: Sequence[Mapping[str, str]]) -> list[str]:
notes: list[str] = []
by_label: dict[tuple[str, str], set[str]] = defaultdict(set)
by_version: dict[str, set[tuple[str, str]]] = defaultdict(set)
for row in rows:
label = (row["protocol_name"], row["series_description"])
version = row["sequence_version"]
if version == NA:
continue
by_label[label].add(version)
by_version[version].add(label)
concealed = [label for label, versions in by_label.items() if len(versions) > 1]
shared = [version for version, labels in by_version.items() if len(labels) > 1]
if concealed:
notes.append(
f"{len(concealed)} protocol/series label combination(s) conceal acquisition-parameter differences."
)
if shared:
notes.append(
f"{len(shared)} acquisition version(s) occur under multiple protocol/series labels."
)
if not notes:
notes.append("Protocol/series labels and acquisition-parameter fingerprints are one-to-one.")
return notes
def write_report(
output_path: Path,
rows: Sequence[dict[str, str]],
source_rows: Sequence[dict[str, str]],
summary: Sequence[dict[str, str]],
inventory: Sequence[dict[str, str]],
review: Sequence[dict[str, str]],
omission_summary: Sequence[dict[str, str]],
unreadable_series: int,
unparsed_series: int,
duplicate_resources: int,
incomplete_copy_groups: int,
) -> None:
participants = {row["participant_id"] for row in inventory}
expected_count = sum(row["nm_expected_from_design"] == "yes" for row in inventory)
present_count = sum(row["nm_present"] == "yes" for row in inventory)
changed = differing_fields(summary)
table_fields = changed or [
"repetition_time_ms",
"echo_time_ms",
"flip_angle_deg",
"pixel_spacing_row_mm",
"pixel_spacing_col_mm",
"slice_thickness_mm",
"number_of_slices",
]
relationships = "\n".join(f"- {note}" for note in label_parameter_relationships(rows))
changed_text = ", ".join(changed) if changed else "none among fingerprinted fields"
no_counterpart = sum(
row.get("bids_counterpart_status") == "no_bids_counterpart" for row in source_rows
)
indeterminate = sum(
row.get("bids_counterpart_status", "").startswith("match_indeterminate")
or row.get("bids_counterpart_status") in {
"bids_scan_inventory_unavailable",
"bids_session_absent",
"session_unparseable",
}
for row in source_rows
)
signature_candidates = sum(
"acquisition_signature" in row.get("review_reason", "") for row in review
)
xa60_candidates = sum(
"XA60" in row.get("software_versions", "") for row in review
)
text = f"""# Night Owls neuromelanin acquisition audit
## Purpose
This audit inventories unreleased neuromelanin MRI scans from the Night Owls project in preparation for possible future BIDS conversion and OpenNeuro release. It does not modify the raw source archive or current BIDS dataset.
## Data sources
The audit examined the local NOSC sourcedata archive at the series-header level and cross-checked it against both the repository's BIDS participant/session directories and its `scans.tsv` inventories. Dates and times are used transiently to link raw series to converted BIDS entries, but raw paths, DICOM identifiers, dates, times, and patient fields are deliberately omitted from every output.
Confirmed-series detection requires a case-insensitive match for `neuromelanin`, a punctuation/spacing variant of `neuro-melanin`, `nmri`, or `substantia nigra` in `ProtocolName`, `SeriesDescription`, or the source series name. Independently, every readable series is scored against the observed NM acquisition signature (approximately TR 641 ms, TE 3.97 ms, flip angle 50 degrees, 0.64 mm in-plane resolution, 2D gradient echo, six repeats, and 24 slices). High-scoring series, generic `NM` labels supported by metadata, and T2-star/MT-like labels are surfaced for review but not silently promoted to confirmed NM.
## Overall inventory
- Participants examined: {len(participants)}
- Study sessions in the union of source and BIDS inventories: {len(inventory)}
- Sessions expected to contain NM from the even-session design: {expected_count}
- Sessions actually containing NM: {present_count}
- Confirmed NM series: {len(rows)}
- Unique readable acquisition versions: {len(summary)}
- Plausible non-confirmed series requiring manual review: {len(review)}
- Name-independent high-signature candidates: {signature_candidates}
- Review candidates whose software reports XA60: {xa60_candidates}
- Deduplicated raw series reconciled: {len(source_rows)}
- Raw series with no time-matched BIDS scan entry: {no_counterpart}
- Raw/BIDS matches that remain indeterminate: {indeterminate}
## Acquisition versions
Version labels are assigned deterministically by sorting normalized acquisition fingerprints; subject/session identifiers, labels, scanner/software metadata, UIDs, paths, dates, and times are excluded from the fingerprint. The fingerprint parameters that differ across versions are: {changed_text}.
{markdown_table(summary, table_fields) if summary else "No readable confirmed NM series were available for version comparison."}
Label-versus-parameter checks:
{relationships}
## Completeness and problems
- Expected even sessions with no confirmed NM data: {status_sessions(inventory, "expected_missing")}
- Odd sessions with confirmed NM data: {status_sessions(inventory, "unexpected_found")}
- Sessions with multiple confirmed NM series: {status_sessions(inventory, "multiple_nm_series")}
- BIDS sessions absent from source: {status_sessions(inventory, "source_missing")}
- Source sessions absent from BIDS: {status_sessions(inventory, "bids_session_missing")}
- Raw series with no readable representative DICOM header: {unreadable_series}
- Readable series without a parseable participant/session path: {unparsed_series}
- Duplicate source-resource copies excluded from series counts: {duplicate_resources}
- Series with a less-complete duplicate source copy: {incomplete_copy_groups}
- Plausible but uncertain candidate series: {len(review)} (see `neuromelanin_candidate_review.tsv`)
- Recurrent unmatched source-series patterns: {len(omission_summary)} (see `source_to_bids_omission_summary.tsv`)
## Source-to-BIDS reconciliation
`source_to_bids_series_reconciliation.tsv` contains one PHI-safe row per deduplicated raw series. A series is `converted_or_listed` when its private source acquisition clock matches one or more entries in that session's deidentified BIDS `scans.tsv` within two seconds. `no_bids_counterpart` means a BIDS scan inventory exists for the session but no such time match was found. Missing scan inventories or source times are reported as indeterminate, not as omissions.
`source_to_bids_omission_summary.tsv` groups only `no_bids_counterpart` rows by scanner software, labels, and acquisition geometry, then reports how often each pattern recurs and in which public subject/session identifiers. Localizers and other recognizable setup scans are flagged as likely intentional exclusions rather than removed from the table. This is the primary table for finding a consistently skipped XA60 or differently named NM protocol.
## Interpretation
The sequence versions are defined by normalized, scientifically meaningful acquisition parameters rather than protocol labels. Parameters that vary are shown in the compact table above; complete acquisition metadata and session membership are in `neuromelanin_sequence_summary.tsv`. Scanner model and software version are reported but do not independently define versions. Confirmed counts remain label-conservative; the candidate and reconciliation tables are required for the completeness conclusion.
## BIDS implications — preliminary only
The historical HeuDiConv heuristic maps any protocol containing `neuromelanin` to `_acq-NM_T2star`. Phase 2 should verify the appropriate suffix, entities, and required metadata against the current BIDS specification for each empirical acquisition variant. It should also determine how magnetization-transfer preparation is represented, whether separate `acq-` labels are warranted, and how to handle missing, unexpected, or repeated series before conversion. No BIDS or OpenNeuro data were changed by this audit.
"""
output_path.write_text(text, encoding="utf-8")
def validate_outputs(
rows: Sequence[dict[str, str]],
summary: Sequence[dict[str, str]],
inventory: Sequence[dict[str, str]],
) -> None:
if len({id(row) for row in rows}) != len(rows):
raise RuntimeError("Internal manifest row duplication detected")
summary_count = sum(int(row["n_series"]) for row in summary)
readable_count = sum(row["sequence_version"] != NA for row in rows)
if summary_count != readable_count:
raise RuntimeError("Sequence summary counts do not reconcile with manifest")
inventory_count = sum(int(row["n_nm_series"]) for row in inventory)
if inventory_count != len(rows):
raise RuntimeError("Session inventory counts do not reconcile with manifest")
def privacy_check(output_dir: Path) -> None:
forbidden = [
re.compile(r"(?:Study|Series|SOP)InstanceUID", re.IGNORECASE),
re.compile(r"Patient(?:Name|Birth|Address|ID)", re.IGNORECASE),
re.compile(r"Acquisition(?:Date|Time|DateTime)", re.IGNORECASE),
re.compile(r"/Users/|/home/|/ZPOOL/|[A-Za-z]:\\\\"),
re.compile(r"\b\d+(?:\.\d+){3,}\b"),
re.compile(r"\b(?:19|20)\d{6}\b"),
re.compile(r"\b(?:[01]\d|2[0-3])[0-5]\d[0-5]\d(?:\.\d+)?\b"),
]
for path in sorted(output_dir.glob("*")):
if path.suffix not in {".tsv", ".md"}:
continue
text = path.read_text(encoding="utf-8")
for pattern in forbidden:
if pattern.search(text):
raise RuntimeError(
f"Privacy check rejected {path.name}: matched forbidden pattern {pattern.pattern!r}"
)
def parse_args(argv: Sequence[str] | None = None) -> argparse.Namespace:
parser = argparse.ArgumentParser(description=__doc__)
parser.add_argument("--source-root", type=Path, required=True, help="Raw NOSC source root")
parser.add_argument("--bids-root", type=Path, required=True, help="Current BIDS dataset root")
parser.add_argument("--output-dir", type=Path, required=True, help="Directory for audit outputs")
parser.add_argument("--verbose", action="store_true", help="Enable progress messages")
return parser.parse_args(argv)
def main(argv: Sequence[str] | None = None) -> int:
args = parse_args(argv)
logging.basicConfig(
level=logging.INFO if args.verbose else logging.WARNING,
format="%(levelname)s: %(message)s",
)
if pydicom is None:
LOG.error("pydicom is required; install code/requirements-neuromelanin-audit.txt")
return 2
source_root = args.source_root.resolve()
bids_root = args.bids_root.resolve()
output_dir = args.output_dir.resolve()
if not source_root.is_dir():
LOG.error("Source root is not a readable directory")
return 2
try:
source_rows, source_sessions, unreadable_series, unparsed_series = audit_series(source_root)
source_rows, duplicate_resources, incomplete_copy_groups = deduplicate_resource_copies(
source_rows
)
bids_sessions = discover_bids_sessions(bids_root)
bids_index = discover_bids_scan_index(bids_root)
reconcile_source_to_bids(source_rows, bids_sessions, bids_index)
rows, review = build_candidate_sets(source_rows)
assign_versions(rows)
summary = build_summary(rows)
inventory = build_inventory(rows, source_sessions, bids_sessions)
omission_summary = build_omission_summary(source_rows)
validate_outputs(rows, summary, inventory)
output_dir.mkdir(parents=True, exist_ok=True)
write_tsv(output_dir / "neuromelanin_scan_manifest.tsv", rows, MANIFEST_FIELDS)
write_tsv(output_dir / "neuromelanin_sequence_summary.tsv", summary, SUMMARY_FIELDS)
write_tsv(output_dir / "neuromelanin_session_inventory.tsv", inventory, INVENTORY_FIELDS)
write_tsv(output_dir / "neuromelanin_candidate_review.tsv", review, REVIEW_FIELDS)
write_tsv(
output_dir / "source_to_bids_series_reconciliation.tsv",
source_rows,
RECONCILIATION_FIELDS,
)
write_tsv(
output_dir / "source_to_bids_omission_summary.tsv",
omission_summary,
OMISSION_SUMMARY_FIELDS,
)
write_report(
output_dir / "README.md",
rows,
source_rows,
summary,
inventory,
review,
omission_summary,
unreadable_series,
unparsed_series,
duplicate_resources,
incomplete_copy_groups,
)
privacy_check(output_dir)
except (OSError, RuntimeError) as error:
LOG.error("%s", error)
return 1
LOG.info(
"Wrote %d confirmed series, %d review candidates, %d versions, %d sessions, and %d source/BIDS rows",
len(rows),
len(review),
len(summary),
len(inventory),
len(source_rows),
)
return 0
if __name__ == "__main__":
sys.exit(main())

audit_neuromelanin.py at commit ec5197e, under MIT · at the source

Overview

Authors: Matthew Mattoni¹, Shenghan Wang¹, Cooper J Sharp¹, Thomas M Olino¹, David V Smith¹

ORCID iDs: Matthew Mattoni, Thomas M Olino

Department of Psychology and Neuroscience, Temple University, Philadelphia, Pennsylvania, USA

Institutions: Temple University (United States)

Journal: Human brain mapping, volume 47, issue 5, article e70512

Dates: received 20 October 2025; accepted 11 March 2026; published online 24 March 2026; in print April 2026

Type: Research article · Language: English

License: CC BY-NC-ND

Identifiers: DOI 10.1002/hbm.70512 · PMID 41877495 · PMCID PMC13140683 · OpenAlex W7140217718

Open access: gold, a free copy (OpenAlex)

Status: code verified

Categories: fMRI (modality), human (organism), methods / tools (subfield)

Methods: Connectivity, Preprocessing, fMRI & imaging, Statistics

MeSH: Brain*, Brain Mapping*, Individuality*, Magnetic Resonance Imaging*, Reward*, Adult, Affect, Female, Humans, Image Processing, Computer-Assisted, Male, Reproducibility of Results, Young Adult (* major topic)

Topic: Functional Brain Connectivity Studies (Cognitive Neuroscience, Neuroscience), according to OpenAlex

Funding: National Institute of Mental Health (F31MH134533); NIH HHS (R01AG067011); NIMH NIH HHS (F31MH134533); National Institutes of Health (R01AG067011)

Citations: cited by 1 paper (Europe PMC); 70 references in the paper

Research resources: which is based on Nipype 1.7.0 RRID:SCR_002502

Abstract

The abstract is not reproduced here: the paper's license (CC BY-NC-ND) does not allow it. Read it in the paper, at the publisher or on Europe PMC.

Repository

Its files are read in the Code ↔ Paper reader above, with 7 matches between paragraphs and lines of code.

DVS-Lab/night-owls

License: MIT

State: the link answers, verified on 28 September 2026

Evidence: files inventoried

Commit: ec5197ed1dbf74717497ce69452974f096246dc7, 9 September 2026

Languages: Shell (97), Python (36), MATLAB (20), R (8), Jupyter (1)

Size: 1,838 files, 162 scripts

Software Heritage: not archived

Found in: the text, “Methods”

Holds: README, license file, environment (code/requirements-neuromelanin-audit.txt), tests, 1 notebook

Not found: CITATION.cff, continuous integration, documentation

Tools: FSL (62 files), fMRIPrep (46 files), NumPy (19 files), pandas (14 files), PsychoPy (12 files), Psychtoolbox (12 files), NiBabel (8 files), tidyverse (7 files), Matplotlib (6 files), ANTs (5 files), psych (4 files), lme4 (3 files), lmerTest (3 files), AFNI (2 files), easystats (2 files), ggplot2 (2 files), patchwork (2 files), pydicom (2 files), dcm2niix (1 file), HeuDiConv (1 file)

Availability: 1 check, the latest on 28 September 2026: the link answers

28 September 2026: the link answers

164 files

code/BIDSto3col.sh, Shell, 307 lines
code/L1stats-loop.sh, Shell, 155 lines
code/L1statsSingleTrial-mid.sh, Shell, 99 lines
code/L1statsSingleTrial-mid_FLOBS.sh, Shell, 90 lines
code/L1statsSingleTrial-sharedreward.sh, Shell, 114 lines
code/L1statsSingleTrial-sharedreward_FLOBS.sh, Shell, 98 lines
code/L1stats_FIR.sh, Shell, 129 lines
code/L1stats_FLOBS.sh, Shell, 129 lines
code/L2stats-hpc-subj-qc.sh, Shell, 232 lines
code/L2stats-hpc-subj.sh, Shell, 169 lines
code/L2stats-hpc-t1w.sh, Shell, 114 lines
code/L2stats-hpc.sh, Shell, 132 lines
code/L3-trend.sh, Shell, 167 lines
code/L3-ttest.sh, Shell, 227 lines
code/MakeConfounds.py, Python, 62 lines
code/MakeSingleTrialsEV_MID.m, MATLAB, 61 lines
code/MakeSingleTrialsEV_SR.m, MATLAB, 84 lines
code/Mattoni2025/InternalConsistency.R, R, 483 lines
code/Mattoni2025/InternalConsistency_RelDif.R, R, 403 lines
code/Mattoni2025/Intraindividual.R, R, 188 lines
code/Mattoni2025/MoodRatings.R, R, 97 lines
code/Mattoni2025/Organize_BehavData.R, R, 99 lines
code/Mattoni2025/RetestReliability.R, R, 235 lines
code/addIntendedFor_fieldmap-hpc.py, Python, 68 lines
code/addUnits_func.py, Python, 32 lines
code/audit_neuromelanin.py, Python, 1,677 lines, 2 matches
code/check_FIR_outputs.py, Python, 212 lines
code/convertSharedReward2BIDSevents.m, MATLAB, 117 lines
code/count_lss_completion.sh, Shell, 104 lines
code/downloadXNAT.py, Python, 45 lines
code/events_generation.R, R, 188 lines
code/extractData-LSS.sh, Shell, 167 lines
code/extractData-cope.sh, Shell, 180 lines
code/extractData.sh, Shell, 183 lines
code/extractData_L1_FLOBS_EVmetrics.py, Python, 633 lines
code/extractData_LSS_FLOBS.py, Python, 548 lines
code/extractSmoothness.sh, Shell, 148 lines
code/extract_event_metrics_raw_and_fir.py, Python, 546 lines
code/extract_fir_roi_means.py, Python, 236 lines
code/fmriprep-anat/fmriprep_sub-101_ses-01.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-101_ses-02.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-101_ses-03.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-101_ses-06.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-101_ses-07.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-101_ses-08.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-101_ses-09.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-101_ses-10.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-101_ses-11.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-103_ses-01.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-103_ses-02.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-103_ses-03.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-103_ses-04.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-103_ses-05.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-103_ses-06.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-103_ses-07.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-103_ses-08.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-103_ses-09.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-103_ses-10.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-103_ses-11.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-103_ses-12.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-104_ses-01.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-104_ses-02.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-104_ses-03.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-104_ses-04.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-104_ses-05.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-104_ses-06.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-104_ses-07.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-104_ses-08.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-104_ses-09.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-104_ses-10.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-104_ses-11.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-104_ses-12.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-105_ses-01.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-105_ses-02.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-105_ses-03.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-105_ses-04.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-105_ses-05.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-105_ses-06.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-105_ses-07.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-105_ses-08.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-105_ses-09.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-105_ses-10.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-105_ses-11.sh, Shell, 43 lines
code/fmriprep-anat/fmriprep_sub-105_ses-12.sh, Shell, 43 lines
code/fmriprep-hpc-anat.sh, Shell, 63 lines
code/fmriprepOrganize.sh, Shell, 46 lines
code/gen3colfiles.sh, Shell, 32 lines
code/genTedanaMultiSes.py, Python, 89 lines
code/gen_MRIQC-outputs.py, Python, 97 lines
code/gen_MRIQC_anat.py, Python, 86 lines
code/gen_fmriprep_anat.sh, Shell, 107 lines
code/heuristics_XA30.py, Python, 66 lines
code/imageSimilarity.py, Python, 56 lines
code/make_zero_duration.sh, Shell, 48 lines
code/mriqc-hpc.sh, Shell, 79 lines
code/normEcho2.sh, Shell, 101 lines
code/plot_fir_timecourses.py, Python, 108 lines
code/prepdata.sh, Shell, 69 lines
code/resample_sub-101_ses-03_task-mid.sh, Shell, 98 lines
code/run_L1stats-qsub.sh, Shell, 19 lines
code/run_L1stats-qsub_FIR.sh, Shell, 23 lines
code/run_L1stats-qsub_FLOBS.sh, Shell, 23 lines
code/run_L1stats-qsub_LSS-FLOBS.sh, Shell, 31 lines
code/run_L1stats-qsub_LSS.sh, Shell, 31 lines
code/run_L2stats-hpc.sh, Shell, 30 lines
code/run_L3stats.sh, Shell, 27 lines
code/run_fmriprep-hpc-anat.sh, Shell, 19 lines
code/run_fmriprep-qsub.sh, Shell, 11 lines
code/run_gen3colfiles.sh, Shell, 25 lines
code/run_imageSimilarity.sh, Shell, 17 lines
code/run_mriqc-hpc.sh, Shell, 18 lines
code/run_prepdata.sh, Shell, 20 lines
code/run_smoothing-qsub.sh, Shell, 19 lines
code/run_tedana-hpc.sh, Shell, 18 lines
code/run_warpkit-hpc.sh, Shell, 22 lines
code/shiftdates.py, Python, 53 lines
code/smooth-3dBlurToFWHM.sh, Shell, 49 lines
code/striatum_MNItoT1w.sh, Shell, 118 lines
code/sub-103_mriqc.sh, Shell, 53 lines
code/sub-103_ses-02_replaceRun2.sh, Shell, 21 lines
code/tedana-hpc.sh, Shell, 91 lines
code/tests/test_audit_neuromelanin.py, Python, 337 lines
code/vs_timecourse-mid.py, Python, 329 lines
code/vs_timecourse-sr.py, Python, 189 lines
code/vs_timecourse_FIR.py, Python, 350 lines
code/vs_timecourse_unified.py, Python, 828 lines
code/vs_timecourse_unified_wu.py, Python, 757 lines
code/warpkit-hpc.sh, Shell, 124 lines
masks/resample_to_study_mni_grid.sh, Shell, 93 lines, 1 match
stimuli/MID_5button.py, Python, 794 lines, 1 match
stimuli/MID_practice.py, Python, 726 lines, 1 match
stimuli/MoodInduction_5button.py, Python, 304 lines
stimuli/MoodRatings_5button.py, Python, 195 lines
stimuli/SharedReward_5button.py, Python, 444 lines, 1 match
stimuli/SharedReward_practice.py, Python, 399 lines
stimuli/mid/scripts/KbWait2.m, MATLAB, 149 lines
stimuli/mid/scripts/SRM_test.m, MATLAB, 218 lines
stimuli/mid/scripts/generate_MIDtrials.R, R, 42 lines
stimuli/mid/scripts/getBoxNumber.m, MATLAB, 18 lines
stimuli/mid/scripts/getKey.m, MATLAB, 37 lines
stimuli/mid/scripts/getKeyboardNumber.m, MATLAB, 12 lines
stimuli/mid/scripts/recordKeys.m, MATLAB, 111 lines
stimuli/mid/scripts/recordKeysNoBT.m, MATLAB, 117 lines
stimuli/mid/scripts/ret_listfile.m, MATLAB, 32 lines
stimuli/mid/scripts/screenscratch.m, MATLAB, 43 lines
stimuli/mid/scripts/set_MID_threshold.m, MATLAB, 36 lines
stimuli/mid/scripts/waitForKeys.m, MATLAB, 113 lines
stimuli/mid/zOld/Data Exploration.ipynb, Jupyter, 493 lines
stimuli/mid/zOld/MID_ext.py, Python, 612 lines
stimuli/mid/zOld/MID_prac.py, Python, 675 lines
stimuli/mid/zOld/mid.m, MATLAB, 241 lines
stimuli/mid/zOld/mid_noloss.m, MATLAB, 277 lines
stimuli/mid/zOld/mid_noloss_practice.m, MATLAB, 307 lines
stimuli/mid/zOld/mid_practice.m, MATLAB, 305 lines
stimuli/mid/zOld/timing/mid_nosc_rungeneration.m, MATLAB, 56 lines
stimuli/ratings_prac.py, Python, 147 lines
stimuli/rest_neuromelanin.py, Python, 50 lines
stimuli/sharedreward/Alternate_Button_Box/SharedReward_4button.py, Python, 389 lines
stimuli/sharedreward/button_test/button_box_test.py, Python, 89 lines
stimuli/sharedreward/genTrialList.m, MATLAB, 131 lines, 1 match
stimuli/sharedreward/neurodesign/ndtest.py, Python, 73 lines
stimuli/sharedreward/neurodesign/optseq2_cmd.sh, Shell, 27 lines
LICENSE, License, 21 lines
README.md, Text, 71 lines

At the source: github.com/DVS-Lab/night-owls

Tracing map

Proposed by the machine: these links were found in the paper and verified at the source, without human review. The map will receive a Zenodo DOI once one of the paper's authors has validated it with their ORCID.

What the map holds:

1 repository of the authors' code, each at its verified commit, with its license and how the link was found in the paper;
162 scripts, each with its path and the digest of its content;
7 matches between paragraphs of the paper and lines of the code (method lexical-v1);
neither the text of the paper nor the code itself.

Its JSON (tracing-map.json) is deposited on Zenodo with its DOI once the map is validated.

Data

Datasets cited

doi:10.18112/openneuro.ds006707.v1.1.0, at OpenNeuro; found in the references
doi:10.18112/openneuro.ds006707.v1.1.1, at OpenNeuro; found in the text, “Methods”
doi:10.18112/openneuro.ds006707.v1.1.1.3, at OpenNeuro; found in “Data Availability Statement”
doi:10.18112/openneuro.ds006707.v1.1.3, at OpenNeuro; found in the text, “Methods”
neurovault.org/images/775976, at neurovault.org; found in the text, “fMRI Analysis and Moderating Factors”
osf:df5t9, at OSF; found in the text, “Methods”

Data availability statement

The paper has a data availability statement. Its license (CC BY-NC-ND) does not allow reproducing it here; in short, from what the harvester recognized in it:

it points to a dataset: OpenNeuro 10.18112/openneuro.ds006707.v1.1.1.3

Read it in the paper: doi.org/10.1002/hbm.70512.

Versions

The history of this record: each version stored by the harvester or made by a correction of its authors or of the maintainers of its code, and what changed in its facts. The texts of the paper (its abstract, its availability statements) are not part of it; versions that changed only those are not listed.

Version 1, 28 September 2026: the first record

Recorded: type, language, journal, volume, issue, pages, dates, 5 authors, 13 MeSH terms, 4 funders, 68 references, 1 RRID.

Cite

This paper

Mattoni, M., Wang, S., Sharp, C. J., Olino, T. M., & Smith, D. V. (2026). Precision Imaging for Intraindividual Investigation of the Reward Response. Human brain mapping, 47(5), e70512. https://doi.org/10.1002/hbm.70512

BibTeX

@article{mattoni2026precision, author = {Mattoni, Matthew and Wang, Shenghan and Sharp, Cooper J and Olino, Thomas M and Smith, David V}, title = {{Precision Imaging for Intraindividual Investigation of the Reward Response}}, journal = {Human brain mapping}, year = {2026}, month = apr, volume = {47}, number = {5}, pages = {e70512}, publisher = {Wiley}, issn = {1065-9471}, doi = {10.1002/hbm.70512}, url = {https://doi.org/10.1002/hbm.70512}, pmid = {41877495}, pmcid = {PMC13140683} }

RIS

TY - JOUR AU - Mattoni, Matthew AU - Wang, Shenghan AU - Sharp, Cooper J AU - Olino, Thomas M AU - Smith, David V TI - Precision Imaging for Intraindividual Investigation of the Reward Response T2 - Human brain mapping J2 - Hum Brain Mapp PY - 2026 DA - 2026/04/01 VL - 47 IS - 5 SP - e70512 SN - 1065-9471 PB - Wiley DO - 10.1002/hbm.70512 UR - https://doi.org/10.1002/hbm.70512 LA - en ER -

CSL-JSON

{ "id": "10.1002/hbm.70512", "type": "article-journal", "title": "Precision Imaging for Intraindividual Investigation of the Reward Response", "container-title": "Human brain mapping", "author": [ { "family": "Mattoni", "given": "Matthew" }, { "family": "Wang", "given": "Shenghan" }, { "family": "Sharp", "given": "Cooper J" }, { "family": "Olino", "given": "Thomas M" }, { "family": "Smith", "given": "David V" } ], "container-title-short": "Hum Brain Mapp", "volume": "47", "issue": "5", "page": "e70512", "DOI": "10.1002/hbm.70512", "PMID": "41877495", "PMCID": "PMC13140683", "ISSN": "1065-9471", "publisher": "Wiley", "URL": "https://doi.org/10.1002/hbm.70512", "language": "en", "issued": { "date-parts": [ [ 2026, 4, 1 ] ] } }

The tracing map gets a citation of its own once an author has validated it and it has a DOI.

Contribute

The authors of this paper can claim it, correct its record and validate its tracing map, and the maintainers of its code (its owner, or a public member of its organization) correct what it says of their repository; anyone signed in can ask for its removal. Every request goes to OSCR's own machine, which answers it; your account page follows them.

Sign in with ORCID to claim this paper as one of its authors, correct its record or validate its tracing map: when the paper's metadata lists your ORCID iD, you are recognized at once. Maintainers of its code: sign in with GitHub, then claim the repository on your account page.

Request its removal

To ask OSCR to remove this record, the copies of its authors' scripts or its tracing map, use the removal request page: signed in, you say who you are, what to remove and why, then review and confirm the request. Published rules decide every request (how).

Discussion, reproductions, activity

Discussion: questions and error reports about this paper and its code, from signed-in readers and its authors. It opens with sign-in.

Reproductions: reports from readers who ran the authors' code: what they reproduced, with which environment, commit and data. It opens with sign-in.

Activity: what happens around this paper: new versions of its record, its map's validation, discussions and reproductions. It opens with sign-in.