curated_breast_imaging_ddsm

  • 説明

CBIS-DDSM (DDSM の厳選された乳房画像処理サブセット) は、スクリーニング マンモグラフィー用デジタル データベース (DDSM) の更新および標準化されたバージョンです。 DDSM は、2,620 件のスキャンされたフィルム マンモグラフィー検査のデータベースです。これには、正常、良性、悪性の症例と検証された病理情報が含まれています。

デフォルトの設定は、従来の画像分類設定で解決するタスクを組み立てるために、( http://arxiv.org/abs/1708.09427 ) の説明に従って、元のマンモグラムから抽出されたパッチで構成されています。

データセットに含まれる画像をダウンロードして読み取るには特別なソフトウェアとライブラリが必要であるため、TFDS はユーザーが元の DCIM ファイルをダウンロードして PNG に変換していることを前提としています。

再現可能な結果を​​保証するには、次のコマンド (または同等のコマンド) を使用して PNG ファイルを生成する必要があります。

find $DATASET_DCIM_DIR -name '*.dcm' | \
xargs
-n1 -P8 -I{} bash -c 'f={}; dcmj2pnm $f | convert - ${f/.dcm/.png}'

結果のイメージは、 <manual_dir>/Mass-Training_P_01981_RIGHT_MLO_1/1.3.6.../000000.pngのように、 manual_dirに配置する必要があります。

@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/patches (デフォルト設定)

  • 構成の説明: 石灰化と集団の両方の症例と、異常のないパスを含むパッチ。従来の 5 クラス分類タスクとして設計されています。

  • ダウンロードサイズ: 2.01 MiB

  • データセットのサイズ: 801.46 MiB

  • 分割:

スプリット
'test' 9,770
'train' 49,780
'validation' 5,580
  • 機能の構造:
FeaturesDict({
   
'id': Text(shape=(), dtype=string),
   
'image': Image(shape=(None, None, 1), dtype=uint8),
   
'label': ClassLabel(shape=(), dtype=int64, num_classes=5),
})
  • 機能ドキュメント:
特徴クラスDタイプ説明
特徴辞書
ID文章
画像画像(なし、なし、1) uint8
ラベルクラスラベルint64

視覚化

curated_breast_imaging_ddsm/original-calc

  • 構成の説明: ロスレス PNG で圧縮された石灰化症例の元の画像。

  • ダウンロードサイズ: 1.06 MiB

  • データセットのサイズ: 4.42 GiB

  • 分割:

スプリット
'test' 284
'train' 1,227
  • 機能の構造:
FeaturesDict({
   
'abnormalities': Sequence({
       
'assessment': ClassLabel(shape=(), dtype=int64, num_classes=6),
       
'calc_distribution': ClassLabel(shape=(), dtype=int64, num_classes=10),
       
'calc_type': ClassLabel(shape=(), dtype=int64, num_classes=48),
       
'id': int32,
       
'mask': Image(shape=(None, None, 1), dtype=uint8),
       
'pathology': ClassLabel(shape=(), dtype=int64, num_classes=3),
       
'subtlety': ClassLabel(shape=(), dtype=int64, num_classes=6),
   
}),
   
'breast': ClassLabel(shape=(), dtype=int64, num_classes=2),
   
'id': Text(shape=(), dtype=string),
   
'image': Image(shape=(None, None, 1), dtype=uint8),
   
'patient': Text(shape=(), dtype=string),
   
'view': ClassLabel(shape=(), dtype=int64, num_classes=2),
})
  • 機能ドキュメント:
特徴クラスDタイプ説明
特徴辞書
異常順序
異常・評価クラスラベルint64
異常/calc_distributionクラスラベルint64
異常/calc_typeクラスラベルint64
異常/IDテンソルint32
異常/マスク画像(なし、なし、1) uint8
異常/病理クラスラベルint64
異常/微妙クラスラベルint64
クラスラベルint64
ID文章
画像画像(なし、なし、1) uint8
忍耐強い文章
ビュークラスラベルint64

視覚化

curated_breast_imaging_ddsm/オリジナル-マス

  • 構成の説明: ロスレス PNG で圧縮された大量症例の元の画像。

  • ダウンロードサイズ: 966.57 KiB

  • データセットのサイズ: 4.80 GiB

  • 分割:

スプリット
'test' 348
'train' 1,166
  • 機能の構造:
FeaturesDict({
   
'abnormalities': Sequence({
       
'assessment': ClassLabel(shape=(), dtype=int64, num_classes=6),
       
'id': int32,
       
'mask': Image(shape=(None, None, 1), dtype=uint8),
       
'mass_margins': ClassLabel(shape=(), dtype=int64, num_classes=20),
       
'mass_shape': ClassLabel(shape=(), dtype=int64, num_classes=21),
       
'pathology': ClassLabel(shape=(), dtype=int64, num_classes=3),
       
'subtlety': ClassLabel(shape=(), dtype=int64, num_classes=6),
   
}),
   
'breast': ClassLabel(shape=(), dtype=int64, num_classes=2),
   
'id': Text(shape=(), dtype=string),
   
'image': Image(shape=(None, None, 1), dtype=uint8),
   
'patient': Text(shape=(), dtype=string),
   
'view': ClassLabel(shape=(), dtype=int64, num_classes=2),
})
  • 機能ドキュメント:
特徴クラスDタイプ説明
特徴辞書
異常順序
異常・評価クラスラベルint64
異常/IDテンソルint32
異常/マスク画像(なし、なし、1) uint8
異常/質量マージンクラスラベルint64
異常/腫瘤の形状クラスラベルint64
異常/病理クラスラベルint64
異常/微妙クラスラベルint64
クラスラベルint64
ID文章
画像画像(なし、なし、1) uint8
忍耐強い文章
ビュークラスラベルint64

視覚化