IDC CT4Harmonization-Multicentric: Hemangioma, Pathologically Benign, Metastatic disease CT (1 phantom subject): sample notebook

A worked notebook for IDC CT4Harmonization-Multicentric: Hemangioma, Pathologically Benign, Metastatic disease CT (1 phantom subject), with 4 code cells and the output each one printed. On Scigantic it opens in a hosted JupyterLab with the dataset already mounted, so nothing is downloaded first.

Provider
NCI Imaging Data Commons; CT4Harmonization-Multicentric contributors; tables and sample by Scigantic
License
Per series (series.parquet license_short_name): CC BY 4.0 2,756. Cite the source DOI(s) and the IDC.
Size
4 parquet tables plus 1 sample DICOM files (4 MB)
Stored at
Amazon S3, s3://scigantic-idc-open/ct4harmonization_multicentric (us-east-1)
Access
Free and public. No account with the data provider is needed.
DOI
10.7937/m0pb-bh69

The notebook

CT4Harmonization-Multicentric on the Imaging Data Commons: 1 phantom subject, 2,756 series, 246.46 GB

Hemangioma, Pathologically Benign, Metastatic disease imaging of the liver phantom from the NCI Imaging Data Commons (IDC). Series by modality: SEG 1,378, CT 1,378.

This mount is the collection's index plus a real sample, not the collection: series.parquet (one row per DICOM series, with the public S3 location of each), studies.parquet, patients.parquet, and sample/ folders holding one SEG series (CC BY 4.0). Anything else is one anonymous S3 read away. README.md and BUILD_REPORT.json at the mount root say exactly how this was built (idc-index, IDC v25).

The code below reads all of it with the scigantic-nci package (pip install "scigantic-nci[dicom]", already installed in this image): from scigantic_nci import idc, then calls such as idc.series("ct4harmonization_multicentric"), idc.read_sample and idc.pull_series. Run anywhere else, the same calls read s3://scigantic-idc-open/ct4harmonization_multicentric/ anonymously.

import os, json
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import pydicom
from scigantic_nci import idc, NciError
plt.rcParams['figure.dpi'] = 72  # keeps the saved figures small enough for the notebook preview
try:
    from matplotlib_inline.backend_inline import set_matplotlib_formats
    set_matplotlib_formats('jpeg')  # slide renders are photographic; JPEG is a fifth the size of PNG here
except ImportError:
    pass

C = 'ct4harmonization_multicentric'
# scigantic_nci reads the mounted archive when its BUILD_REPORT.json names this collection, else s3://scigantic-idc-open/<collection>/ anonymously.
M = os.environ.get('SCIGANTIC_MOUNT_PATH', '/mnt/archive')
on_mount = os.path.exists(f'{M}/BUILD_REPORT.json') and json.load(open(f'{M}/BUILD_REPORT.json')).get('collection_id') == C
print('reading', C, 'from', M if on_mount else f's3://scigantic-idc-open/{C}/')

report = idc.report(C)
series = idc.series(C)  # one row per DICOM series, with its public S3 location
patients = idc.patients(C)
studies = idc.studies(C)
print(report['collection_name'], '| IDC', report['idc_data_version'], '| built', report['built_at'][:10])
print(f"{len(patients):,} patients, {len(studies):,} studies, {len(series):,} series, {series.series_size_MB.sum()/1024:,.1f} GB in the IDC buckets\n")
by_mod = series.groupby('Modality').agg(series=('SeriesInstanceUID', 'count'), patients=('PatientID', 'nunique'), GB=('series_size_MB', lambda s: round(s.sum() / 1024, 2))).sort_values('series', ascending=False)
print(by_mod.to_string())
fig, ax = plt.subplots(figsize=(7, 3))
by_mod['patients'].plot.bar(ax=ax, color='#4a6fa5'); ax.set_ylabel('patients'); ax.set_title(f"{report['collection_name']}: patients per modality"); plt.tight_layout()

Output:

reading ct4harmonization_multicentric from /mnt/archive
CT4Harmonization-Multicentric | IDC v25 | built 2026-11-17
1 patients, 1,378 studies, 2,756 series, 246.5 GB in the IDC buckets

          series  patients      GB
Modality                          
CT          1378         1  239.76
SEG         1378         1    6.70

1. Studies and patients

A study is one imaging visit (all series acquired together); a patient may have several. studies.parquet collapses the series table to one row per study with the modalities present, patients.parquet to one row per patient.

print('Studies per patient:'); print(patients.n_studies.describe()[['min', '50%', 'max']].to_string())
print('\nModality combinations per study (top 8):'); print(studies.modalities.value_counts().head(8).to_string())
sex = patients.PatientSex.replace('', np.nan).dropna()
if len(sex): print('\nPatientSex:', sex.value_counts().to_dict())
ages = pd.to_numeric(series.PatientAge.dropna().astype(str).str.extract(r'(\d+)')[0], errors='coerce')
if ages.notna().sum() > 10: print(f'PatientAge at imaging: median {ages.median():.0f} y (IQR {ages.quantile(.25):.0f}-{ages.quantile(.75):.0f}) on {ages.notna().sum():,} series')
if 'BodyPartExamined' in series: print('\nBodyPartExamined:', series.BodyPartExamined.dropna().value_counts().head(6).to_dict())
print('\nLicenses (per series):', series.license_short_name.value_counts().to_dict())

Output:

Studies per patient:
min    1378.0
50%    1378.0
max    1378.0

Modality combinations per study (top 8):
modalities
CT, SEG    1378

PatientSex: {'O': 1}

BodyPartExamined: {'ABDOMEN': 1000, 'CHEST': 900, 'CHESTABDOMEN': 856}

Licenses (per series): {'CC BY 4.0': 2756}

2. Pull the smallest CT series from IDC and open it

Nothing is under sample/ on this mount, so the cell below reads the smallest CT series straight from IDC's public bucket into the pod's /tmp (anonymous S3, no credentials). Every row of series.parquet has aws_bucket and crdc_series_uuid; the objects under that prefix are the series' instances, one .dcm per slice. idc.pull_series(row, dest) downloads them and returns the local paths, which pydicom reads below. Uncompressed transfer syntaxes are preferred so the read needs no decoder plugin. Check series_size_MB before pulling a bigger one: a pod has a few GB of local disk.

cand = idc.series(C, modality='CT')  # the modality filter is pushed down to parquet
if 'CT' in ('CT', 'MR', 'PT') and (cand.instanceCount >= 10).any():
    cand = cand[cand.instanceCount >= 10]  # a real stack, not a scout
cand = cand.assign(_compressed=cand.transfer_syntax_name.fillna('').str.contains('JPEG|RLE', regex=True))
row = cand.sort_values(['_compressed', 'series_size_MB']).iloc[0]
print(row.PatientID, row.Modality, int(row.instanceCount), 'instances,', round(idc.series_size(row), 1), 'MB,', row.license_short_name, '->', row.series_aws_url)
print('IDC viewer:', idc.viewer_url(row))
folder = '/tmp/idc_pull/' + row.crdc_series_uuid
files = idc.pull_series(row, folder)  # anonymous read of s3://<aws_bucket>/<crdc_series_uuid>/, folder marker skipped
print(len(files), 'files ->', folder)
# scigantic_nci returns file paths for a pulled series (read_sample only reads the mirrored samples), so pydicom stacks them here
ds = [pydicom.dcmread(f) for f in files]
ds = [d for d in ds if 'PixelData' in d]
try:
    ds[0].pixel_array
except RuntimeError as e:  # compressed transfer syntax with no decoder plugin in this image
    print(f'This series is stored as {ds[0].file_meta.TransferSyntaxUID.name}; this image cannot decode it (pip install "pylibjpeg[all]" adds the decoder). Skipping the render.')
    ds = []
if ds:
    ds.sort(key=lambda d: (float(d.ImagePositionPatient[2]) if 'ImagePositionPatient' in d else 0.0, int(d.get('InstanceNumber', 0))))
    def frame(d):
        a = d.pixel_array
        return a[a.shape[0] // 2] if int(d.get('NumberOfFrames', 1) or 1) > 1 else a  # multi-frame instance (grayscale or RGB): middle frame
    shapes = sorted({frame(d).shape for d in ds})
    if len(shapes) == 1 and len(ds) >= 3:
        vol = np.stack([frame(d) for d in ds]).astype(np.float32)
        vol = vol * float(ds[0].get('RescaleSlope', 1)) + float(ds[0].get('RescaleIntercept', 0))  # Hounsfield units
        lo, hi = (-160, 240) if ds[0].Modality == 'CT' else (np.percentile(vol, 1), np.percentile(vol, 99))
        picks = [(vol[i], f'slice {i + 1}/{len(ds)}') for i in (len(ds) // 4, len(ds) // 2, 3 * len(ds) // 4)]
    else:  # 2D modalities or a series that mixes image sizes (scouts, crops): show instances one by one
        picks = []
        for d in ds[:3]:
            img = frame(d).astype(np.float32) * float(d.get('RescaleSlope', 1)) + float(d.get('RescaleIntercept', 0))
            picks.append((img, f'{d.get("SeriesDescription", "")} {img.shape[1]}x{img.shape[0]}'))
        vol = np.concatenate([p[0].ravel() for p in picks]); lo, hi = np.percentile(vol, 1), np.percentile(vol, 99)
    d0 = ds[0]
    print(f'{d0.Modality} {len(ds)} instances, image shapes {shapes} | spacing {d0.get("PixelSpacing") or d0.get("ImagerPixelSpacing")} mm, slice thickness {d0.get("SliceThickness")} mm | {d0.get("Manufacturer")} {d0.get("ManufacturerModelName")} | {d0.get("SeriesDescription")}')
    fig, axes = plt.subplots(1, len(picks), figsize=(3.6 * len(picks), 3.8))
    for ax, (img, label) in zip(np.atleast_1d(axes), picks):
        ax.imshow(img if img.ndim == 2 else np.clip((img - lo) / max(hi - lo, 1e-6), 0, 1), cmap='gray', vmin=lo, vmax=hi); ax.set_title(label, fontsize=9); ax.axis('off')  # RGB (ultrasound) is scaled to 0-1
    plt.suptitle(f"CT4Harmonization-Multicentric: smallest CT series, patient {row.PatientID} ({row.license_short_name})"); plt.tight_layout()

Output:

CT4Harmonization-Multicentric CT 275 instances, 144.8 MB, CC BY 4.0 -> s3://idc-open-data/5e04a037-4ef9-41e7-9409-7059024929fd/*
IDC viewer: https://viewer.imaging.datacommons.cancer.gov/viewer/1.2.840.113619.2.55.3.313318667.350.16157920.2431755691890547653?seriesInstanceUID=1.2.840.113619.2.55.3.313318667.350.1615792084.33
275 files -> /tmp/idc_pull/5e04a037-4ef9-41e7-9409-7059024929fd
CT 275 instances, image shapes [(512, 512)] | spacing [0.683594, 0.683594] mm, slice thickness 2.500000 mm | GE MEDICAL SYSTEMS BrightSpeed S | F1_Harmonized_6mGy_IR - #2

3. The sample segmentation

sample/SEG_1.2.826.0.1.3680043.8.498.12604332692176444923647811866999541504/ is a DICOM SEG series of patient CT4Harmonization-Multicentric (4.2 MB, CC BY 4.0): a stack of binary frames, one per segment and source slice, with SegmentSequence naming each segment and ReferencedSeriesSequence pointing at the image series it was drawn on. A few collections label ordinary image instances (one label map per slice) as SEG; the cell handles both. idc.sample_files lists the files; on the mount they are read in place, and elsewhere the same series is pulled from IDC with idc.pull_series. The source image series can be pulled from IDC the same way as any other row of series.parquet.

# scigantic_nci has no SEG reader, and read_sample decodes one frame per instance, which pydicom cannot do for some 1-bit SEG objects,
# so the SEG files are read whole with pydicom: in place on the mount (sample_files gives paths relative to the archive root), else pulled from IDC.
seg_sample = idc.samples(C).query("modality == 'SEG'").iloc[0]
if on_mount:
    seg_files = [os.path.join(M, f) for f in idc.sample_files(C, 'SEG')]
else:
    seg_files = idc.pull_series(series[series.SeriesInstanceUID == seg_sample.SeriesInstanceUID].iloc[0], '/tmp/idc_pull/' + seg_sample.SeriesInstanceUID)
seg_ds = [pydicom.dcmread(f) for f in seg_files]
seg = seg_ds[0]
if 'SegmentSequence' in seg:  # a DICOM Segmentation object
    print(seg.Modality, seg.get('SeriesDescription'), '|', int(seg.get('NumberOfFrames', 1)), 'frames of', seg.Rows, 'x', seg.Columns, '|', seg.get('ContentCreatorName'))
    print('segments:', [(int(x.SegmentNumber), str(x.SegmentLabel), str(x.get('SegmentAlgorithmType'))) for x in seg.SegmentSequence][:8])
    mask = seg.pixel_array
else:  # labelled SEG but stored as ordinary images: one label map per slice
    seg_ds.sort(key=lambda d: (float(d.ImagePositionPatient[2]) if 'ImagePositionPatient' in d else 0.0, int(d.get('InstanceNumber', 0))))
    print(seg.Modality, seg.get('SeriesDescription'), '|', len(seg_ds), seg.SOPClassUID.name, 'instances of', seg.Rows, 'x', seg.Columns, '(label maps, not a Segmentation object)')
    mask = np.stack([d.pixel_array for d in seg_ds])
    values, counts = np.unique(mask, return_counts=True)
    mask = mask != values[counts.argmax()]  # label maps carry a nonzero background value; everything else is labelled
ref = seg.ReferencedSeriesSequence[0].SeriesInstanceUID if 'ReferencedSeriesSequence' in seg else None
src = series[series.SeriesInstanceUID == ref]
if len(src):
    r = src.iloc[0]
    print(f'drawn on the {r.Modality} series of patient {r.PatientID}: {int(r.instanceCount)} instances, {r.series_size_MB:.1f} MB, {r.license_short_name}, {r.series_aws_url}')
mask = (mask[None] if mask.ndim == 2 else mask) > 0
i = int(mask.reshape(mask.shape[0], -1).sum(1).argmax())  # the frame with the most labelled pixels
print(f'{int(mask.sum()):,} labelled pixels over {mask.shape[0]} frames; densest frame {i + 1}')
fig, ax = plt.subplots(figsize=(3.8, 3.8)); ax.imshow(mask[i], cmap='gray'); ax.axis('off'); ax.set_title(f"CT4Harmonization-Multicentric: sample SEG, frame {i + 1}", fontsize=9); plt.tight_layout()

Output:

SEG Manually Segmented Liver ROIs | 126 frames of 512 x 512 | NIfTI to SEG
segments: [(1, 'normal1', 'MANUAL'), (2, 'normal2', 'MANUAL'), (3, 'cyst1', 'MANUAL'), (4, 'cyst2', 'MANUAL'), (5, 'hemangioma', 'MANUAL'), (6, 'metastatsis', 'MANUAL')]
drawn on the CT series of patient CT4Harmonization-Multicentric: 343 instances, 180.7 MB, CC BY 4.0, s3://idc-open-data/78f01751-24e5-4d18-b0f0-e8278875e0bd/*
241,912 labelled pixels over 126 frames; densest frame 39

Try this in the AI assistant

The chat panel is prefilled with the first of these.

Licenses are per series (license_short_name in series.parquet). Cite the collection DOI(s) in README.md and the IDC papers (Fedorov et al. 2021, 2023) when you publish.

Run it

Open this notebook to run it, see the dataset page for the full description, or browse the full catalog.