Hình ảnh: Khám phá trong Know dữ liệu của bạn
Mô tả:
CBIS-DDSM (Bộ phân định hình ảnh vú được sắp xếp của DDSM) là phiên bản cập nhật và tiêu chuẩn hóa của Cơ sở dữ liệu kỹ thuật số cho chụp nhũ ảnh tầm soát (DDSM). DDSM là cơ sở dữ liệu của 2.620 nghiên cứu chụp nhũ ảnh bằng phim được quét. Nó chứa các trường hợp bình thường, lành tính và ác tính với thông tin bệnh lý đã được xác minh.
Các cấu hình mặc định được làm bằng các bản vá lỗi được chiết xuất từ các nhũ ảnh gốc, theo mô tả từ http://arxiv.org/abs/1708.09427, để khung nhiệm vụ để giải quyết trong một khung cảnh phân loại hình ảnh truyền thống.
Trang chủ: https://wiki.cancerimagingarchive.net/display/Public/CBIS-DDSM
Source code:
tfds.image_classification.CuratedBreastImagingDDSM
phiên bản:
-
2.0.1
: New split API ( https://tensorflow.org/datasets/splits ) -
3.0.0
(mặc định): lấy mẫu cắt xén Better ( https://github.com/tensorflow/datasets/pull/2502 )
-
Hướng dẫn tải về Hướng dẫn sử dụng: Bộ dữ liệu này đòi hỏi bạn phải tải về dữ liệu nguồn bằng tay vào
download_config.manual_dir
(mặc định là~/tensorflow_datasets/downloads/manual/
):
Bạn có thể tải hình ảnh từ https://wiki.cancerimagingarchive.net/display/Public/CBIS-DDSM
Vì cần có phần mềm và thư viện đặc biệt để tải xuống và đọc các hình ảnh có trong tập dữ liệu, TFDS giả định rằng người dùng đã tải xuống các tệp DCIM gốc và chuyển đổi chúng sang PNG.
Các lệnh sau (hoặc tương đương) nên được sử dụng để tạo tệp PNG, nhằm đảm bảo kết quả có thể lặp lại:
find $DATASET_DCIM_DIR -name '*.dcm' | \
xargs -n1 -P8 -I{} bash -c 'f={}; dcmj2pnm $f | convert - ${f/.dcm/.png}'
Kết quả hình ảnh nên được đặt trong manual_dir
, như: <manual_dir>/Mass-Training_P_01981_RIGHT_MLO_1/1.3.6.../000000.png
.
Tự động lưu trữ ( tài liệu ): Không
Phím giám sát (Xem
as_supervised
doc ):None
Trích dẫn:
@misc{CBIS_DDSM_Citation,
doi = {10.7937/k9/tcia.2016.7o02s9cy},
url = {https://wiki.cancerimagingarchive.net/x/lZNXAQ},
author = {Sawyer-Lee, Rebecca and Gimenez, Francisco and Hoogi, Assaf and Rubin, Daniel},
title = {Curated Breast Imaging Subset of DDSM},
publisher = {The Cancer Imaging Archive},
year = {2016},
}
@article{TCIA_Citation,
author = {
K. Clark and B. Vendt and K. Smith and J. Freymann and J. Kirby and
P. Koppel and S. Moore and S. Phillips and D. Maffitt and M. Pringle and
L. Tarbox and F. Prior
},
title = { {The Cancer Imaging Archive (TCIA): Maintaining and Operating a
Public Information Repository} },
journal = {Journal of Digital Imaging},
volume = {26},
month = {December},
year = {2013},
pages = {1045-1057},
}
@article{DBLP:journals/corr/abs-1708-09427,
author = {Li Shen},
title = {End-to-end Training for Whole Image Breast Cancer Diagnosis using
An All Convolutional Design},
journal = {CoRR},
volume = {abs/1708.09427},
year = {2017},
url = {http://arxiv.org/abs/1708.09427},
archivePrefix = {arXiv},
eprint = {1708.09427},
timestamp = {Mon, 13 Aug 2018 16:48:35 +0200},
biburl = {https://dblp.org/rec/bib/journals/corr/abs-1708-09427},
bibsource = {dblp computer science bibliography, https://dblp.org}
}
curated_breast_imaging_ddsm / patch (cấu hình mặc định)
Config mô tả: Patches chứa cả calsification và các trường hợp quần chúng, cộng với pathces không có bất thường. Được thiết kế như một nhiệm vụ phân loại 5 lớp truyền thống.
Dung lượng tải về:
2.01 MiB
Dataset kích thước:
801.46 MiB
tách:
Tách ra | Các ví dụ |
---|---|
'test' | 9,770 |
'train' | 49.780 |
'validation' | 5.580 |
- Các tính năng:
FeaturesDict({
'id': Text(shape=(), dtype=tf.string),
'image': Image(shape=(None, None, 1), dtype=tf.uint8),
'label': ClassLabel(shape=(), dtype=tf.int64, num_classes=5),
})
- Hình ( tfds.show_examples ):
- Ví dụ ( tfds.as_dataframe ):
curated_breast_imaging_ddsm / original-calc
Config mô tả: hình ảnh gốc trong các trường hợp vôi hóa nén PNG lossless.
Dung lượng tải về:
1.06 MiB
Kích thước tập dữ liệu:
4.42 GiB
tách:
Tách ra | Các ví dụ |
---|---|
'test' | 284 |
'train' | 1.227 |
- Các tính năng:
FeaturesDict({
'abnormalities': Sequence({
'assessment': ClassLabel(shape=(), dtype=tf.int64, num_classes=6),
'calc_distribution': ClassLabel(shape=(), dtype=tf.int64, num_classes=10),
'calc_type': ClassLabel(shape=(), dtype=tf.int64, num_classes=48),
'id': tf.int32,
'mask': Image(shape=(None, None, 1), dtype=tf.uint8),
'pathology': ClassLabel(shape=(), dtype=tf.int64, num_classes=3),
'subtlety': ClassLabel(shape=(), dtype=tf.int64, num_classes=6),
}),
'breast': ClassLabel(shape=(), dtype=tf.int64, num_classes=2),
'id': Text(shape=(), dtype=tf.string),
'image': Image(shape=(None, None, 1), dtype=tf.uint8),
'patient': Text(shape=(), dtype=tf.string),
'view': ClassLabel(shape=(), dtype=tf.int64, num_classes=2),
})
- Hình ( tfds.show_examples ):
- Ví dụ ( tfds.as_dataframe ):
curated_breast_imaging_ddsm / original-mass
Config mô tả: hình ảnh gốc trong các trường hợp khối lượng nén PNG lossless.
Dung lượng tải về:
966.57 KiB
Kích thước tập dữ liệu:
4.80 GiB
tách:
Tách ra | Các ví dụ |
---|---|
'test' | 348 |
'train' | 1.166 |
- Các tính năng:
FeaturesDict({
'abnormalities': Sequence({
'assessment': ClassLabel(shape=(), dtype=tf.int64, num_classes=6),
'id': tf.int32,
'mask': Image(shape=(None, None, 1), dtype=tf.uint8),
'mass_margins': ClassLabel(shape=(), dtype=tf.int64, num_classes=20),
'mass_shape': ClassLabel(shape=(), dtype=tf.int64, num_classes=21),
'pathology': ClassLabel(shape=(), dtype=tf.int64, num_classes=3),
'subtlety': ClassLabel(shape=(), dtype=tf.int64, num_classes=6),
}),
'breast': ClassLabel(shape=(), dtype=tf.int64, num_classes=2),
'id': Text(shape=(), dtype=tf.string),
'image': Image(shape=(None, None, 1), dtype=tf.uint8),
'patient': Text(shape=(), dtype=tf.string),
'view': ClassLabel(shape=(), dtype=tf.int64, num_classes=2),
})
- Hình ( tfds.show_examples ):
- Ví dụ ( tfds.as_dataframe ):