Skip to content

montgomery

load_data ΒΆ

Load and return the montgomery dataset.

Sourced from http://openi.nlm.nih.gov/imgs/collections/NLM-MontgomeryCXRSet.zip. This method will download the data to local storage if the data has not been previously downloaded.

Parameters:

Name Type Description Default
root_dir Optional[str]

The path to store the downloaded data. When path is not provided, the data will be saved into fastestimator_data under the user's home directory.

 None

Returns:

Type Description
CSVDataset

train_data
Source code in fastestimator\fastestimator\dataset\data\montgomery.py 
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
def load_data(root_dir: Optional[str] = None) -> CSVDataset:
    """Load and return the montgomery dataset.

    Sourced from http://openi.nlm.nih.gov/imgs/collections/NLM-MontgomeryCXRSet.zip. This method will download the data
        to local storage if the data has not been previously downloaded.

    Args:
        root_dir: The path to store the downloaded data. When `path` is not provided, the data will be saved into
            `fastestimator_data` under the user's home directory.

    Returns:
        train_data
    """
    home = str(Path.home())

    if root_dir is None:
        root_dir = os.path.join(home, 'fastestimator_data', 'Montgomery')
    else:
        root_dir = os.path.join(os.path.abspath(root_dir), 'Montgomery')
    os.makedirs(root_dir, exist_ok=True)

    csv_path = os.path.join(root_dir, "montgomery.csv")
    data_compressed_path = os.path.join(root_dir, 'NLM-MontgomeryCXRSet.zip')
    extract_folder_path = os.path.join(root_dir, 'MontgomerySet')

    if not os.path.exists(extract_folder_path):
        # download
        if not os.path.exists(data_compressed_path):
            print("Downloading data to {}".format(root_dir))
            wget.download('http://openi.nlm.nih.gov/imgs/collections/NLM-MontgomeryCXRSet.zip',
                          root_dir,
                          bar=bar_custom)

        # extract
        print("\nExtracting file ...")
        with zipfile.ZipFile(data_compressed_path, 'r') as zip_file:
            # There's some garbage data from macOS in the zip file that gets filtered out here
            zip_file.extractall(root_dir, filter(lambda x: x.startswith("MontgomerySet/"), zip_file.namelist()))

    # glob and generate csv
    if not os.path.exists(csv_path):
        img_list = glob(os.path.join(extract_folder_path, 'CXR_png', '*.png'))
        df = pd.DataFrame(data={'image': img_list})
        df['image'] = df['image'].apply(lambda x: os.path.relpath(x, root_dir))
        df['image'] = df['image'].apply(os.path.normpath)
        df['mask_left'] = df['image'].apply(lambda x: x.replace('CXR_png', os.path.join('ManualMask', 'leftMask')))
        df['mask_right'] = df['image'].apply(lambda x: x.replace('CXR_png', os.path.join('ManualMask', 'rightMask')))
        df.to_csv(csv_path, index=False)

    return CSVDataset(csv_path)