Mass detection in digital breast tomosynthesis: Deep convolutional neural network with transfer learning from mammography
Purpose: Develop a computer-aided detection (CAD) system for masses in digital breast tomosynthesis (DBT) volume using a deep convolutional neural network (DCNN) with transfer learning from mammograms. Methods: A data set containing 2282 digitized film and digital mammograms and 324 DBT volumes were...
Gespeichert in:
| Veröffentlicht in: | Medical physics (Lancaster) Jg. 43; H. 12; S. 6654 - 6666 |
|---|---|
| Hauptverfasser: | , , , , , |
| Format: | Journal Article |
| Sprache: | Englisch |
| Veröffentlicht: |
United States
American Association of Physicists in Medicine
01.12.2016
|
| Schlagworte: | |
| ISSN: | 0094-2405, 2473-4209 |
| Online-Zugang: | Volltext |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| Abstract | Purpose:
Develop a computer-aided detection (CAD) system for masses in digital breast tomosynthesis (DBT) volume using a deep convolutional neural network (DCNN) with transfer learning from mammograms.
Methods:
A data set containing 2282 digitized film and digital mammograms and 324 DBT volumes were collected with IRB approval. The mass of interest on the images was marked by an experienced breast radiologist as reference standard. The data set was partitioned into a training set (2282 mammograms with 2461 masses and 230 DBT views with 228 masses) and an independent test set (94 DBT views with 89 masses). For DCNN training, the region of interest (ROI) containing the mass (true positive) was extracted from each image. False positive (FP) ROIs were identified at prescreening by their previously developed CAD systems. After data augmentation, a total of 45 072 mammographic ROIs and 37 450 DBT ROIs were obtained. Data normalization and reduction of non-uniformity in the ROIs across heterogeneous data was achieved using a background correction method applied to each ROI. A DCNN with four convolutional layers and three fully connected (FC) layers was first trained on the mammography data. Jittering and dropout techniques were used to reduce overfitting. After training with the mammographic ROIs, all weights in the first three convolutional layers were frozen, and only the last convolution layer and the FC layers were randomly initialized again and trained using the DBT training ROIs. The authors compared the performances of two CAD systems for mass detection in DBT: one used the DCNN-based approach and the other used their previously developed feature-based approach for FP reduction. The prescreening stage was identical in both systems, passing the same set of mass candidates to the FP reduction stage. For the feature-based CAD system, 3D clustering and active contour method was used for segmentation; morphological, gray level, and texture features were extracted and merged with a linear discriminant classifier to score the detected masses. For the DCNN-based CAD system, ROIs from five consecutive slices centered at each candidate were passed through the trained DCNN and a mass likelihood score was generated. The performances of the CAD systems were evaluated using free-response ROC curves and the performance difference was analyzed using a non-parametric method.
Results:
Before transfer learning, the DCNN trained only on mammograms with an AUC of 0.99 classified DBT masses with an AUC of 0.81 in the DBT training set. After transfer learning with DBT, the AUC improved to 0.90. For breast-based CAD detection in the test set, the sensitivity for the feature-based and the DCNN-based CAD systems was 83% and 91%, respectively, at 1 FP/DBT volume. The difference between the performances for the two systems was statistically significant (p-value < 0.05).
Conclusions:
The image patterns learned from the mammograms were transferred to the mass detection on DBT slices through the DCNN. This study demonstrated that large data sets collected from mammography are useful for developing new CAD systems for DBT, alleviating the problem and effort of collecting entirely new large data sets for the new modality. |
|---|---|
| AbstractList | Purpose:
Develop a computer-aided detection (CAD) system for masses in digital breast tomosynthesis (DBT) volume using a deep convolutional neural network (DCNN) with transfer learning from mammograms.
Methods:
A data set containing 2282 digitized film and digital mammograms and 324 DBT volumes were collected with IRB approval. The mass of interest on the images was marked by an experienced breast radiologist as reference standard. The data set was partitioned into a training set (2282 mammograms with 2461 masses and 230 DBT views with 228 masses) and an independent test set (94 DBT views with 89 masses). For DCNN training, the region of interest (ROI) containing the mass (true positive) was extracted from each image. False positive (FP) ROIs were identified at prescreening by their previously developed CAD systems. After data augmentation, a total of 45 072 mammographic ROIs and 37 450 DBT ROIs were obtained. Data normalization and reduction of non-uniformity in the ROIs across heterogeneous data was achieved using a background correction method applied to each ROI. A DCNN with four convolutional layers and three fully connected (FC) layers was first trained on the mammography data. Jittering and dropout techniques were used to reduce overfitting. After training with the mammographic ROIs, all weights in the first three convolutional layers were frozen, and only the last convolution layer and the FC layers were randomly initialized again and trained using the DBT training ROIs. The authors compared the performances of two CAD systems for mass detection in DBT: one used the DCNN-based approach and the other used their previously developed feature-based approach for FP reduction. The prescreening stage was identical in both systems, passing the same set of mass candidates to the FP reduction stage. For the feature-based CAD system, 3D clustering and active contour method was used for segmentation; morphological, gray level, and texture features were extracted and merged with a linear discriminant classifier to score the detected masses. For the DCNN-based CAD system, ROIs from five consecutive slices centered at each candidate were passed through the trained DCNN and a mass likelihood score was generated. The performances of the CAD systems were evaluated using free-response ROC curves and the performance difference was analyzed using a non-parametric method.
Results:
Before transfer learning, the DCNN trained only on mammograms with an AUC of 0.99 classified DBT masses with an AUC of 0.81 in the DBT training set. After transfer learning with DBT, the AUC improved to 0.90. For breast-based CAD detection in the test set, the sensitivity for the feature-based and the DCNN-based CAD systems was 83% and 91%, respectively, at 1 FP/DBT volume. The difference between the performances for the two systems was statistically significant (p-value < 0.05).
Conclusions:
The image patterns learned from the mammograms were transferred to the mass detection on DBT slices through the DCNN. This study demonstrated that large data sets collected from mammography are useful for developing new CAD systems for DBT, alleviating the problem and effort of collecting entirely new large data sets for the new modality. Develop a computer-aided detection (CAD) system for masses in digital breast tomosynthesis (DBT) volume using a deep convolutional neural network (DCNN) with transfer learning from mammograms. A data set containing 2282 digitized film and digital mammograms and 324 DBT volumes were collected with IRB approval. The mass of interest on the images was marked by an experienced breast radiologist as reference standard. The data set was partitioned into a training set (2282 mammograms with 2461 masses and 230 DBT views with 228 masses) and an independent test set (94 DBT views with 89 masses). For DCNN training, the region of interest (ROI) containing the mass (true positive) was extracted from each image. False positive (FP) ROIs were identified at prescreening by their previously developed CAD systems. After data augmentation, a total of 45 072 mammographic ROIs and 37 450 DBT ROIs were obtained. Data normalization and reduction of non-uniformity in the ROIs across heterogeneous data was achieved using a background correction method applied to each ROI. A DCNN with four convolutional layers and three fully connected (FC) layers was first trained on the mammography data. Jittering and dropout techniques were used to reduce overfitting. After training with the mammographic ROIs, all weights in the first three convolutional layers were frozen, and only the last convolution layer and the FC layers were randomly initialized again and trained using the DBT training ROIs. The authors compared the performances of two CAD systems for mass detection in DBT: one used the DCNN-based approach and the other used their previously developed feature-based approach for FP reduction. The prescreening stage was identical in both systems, passing the same set of mass candidates to the FP reduction stage. For the feature-based CAD system, 3D clustering and active contour method was used for segmentation; morphological, gray level, and texture features were extracted and merged with a linear discriminant classifier to score the detected masses. For the DCNN-based CAD system, ROIs from five consecutive slices centered at each candidate were passed through the trained DCNN and a mass likelihood score was generated. The performances of the CAD systems were evaluated using free-response ROC curves and the performance difference was analyzed using a non-parametric method. Before transfer learning, the DCNN trained only on mammograms with an AUC of 0.99 classified DBT masses with an AUC of 0.81 in the DBT training set. After transfer learning with DBT, the AUC improved to 0.90. For breast-based CAD detection in the test set, the sensitivity for the feature-based and the DCNN-based CAD systems was 83% and 91%, respectively, at 1 FP/DBT volume. The difference between the performances for the two systems was statistically significant (p-value < 0.05). The image patterns learned from the mammograms were transferred to the mass detection on DBT slices through the DCNN. This study demonstrated that large data sets collected from mammography are useful for developing new CAD systems for DBT, alleviating the problem and effort of collecting entirely new large data sets for the new modality. PURPOSEDevelop a computer-aided detection (CAD) system for masses in digital breast tomosynthesis (DBT) volume using a deep convolutional neural network (DCNN) with transfer learning from mammograms.METHODSA data set containing 2282 digitized film and digital mammograms and 324 DBT volumes were collected with IRB approval. The mass of interest on the images was marked by an experienced breast radiologist as reference standard. The data set was partitioned into a training set (2282 mammograms with 2461 masses and 230 DBT views with 228 masses) and an independent test set (94 DBT views with 89 masses). For DCNN training, the region of interest (ROI) containing the mass (true positive) was extracted from each image. False positive (FP) ROIs were identified at prescreening by their previously developed CAD systems. After data augmentation, a total of 45 072 mammographic ROIs and 37 450 DBT ROIs were obtained. Data normalization and reduction of non-uniformity in the ROIs across heterogeneous data was achieved using a background correction method applied to each ROI. A DCNN with four convolutional layers and three fully connected (FC) layers was first trained on the mammography data. Jittering and dropout techniques were used to reduce overfitting. After training with the mammographic ROIs, all weights in the first three convolutional layers were frozen, and only the last convolution layer and the FC layers were randomly initialized again and trained using the DBT training ROIs. The authors compared the performances of two CAD systems for mass detection in DBT: one used the DCNN-based approach and the other used their previously developed feature-based approach for FP reduction. The prescreening stage was identical in both systems, passing the same set of mass candidates to the FP reduction stage. For the feature-based CAD system, 3D clustering and active contour method was used for segmentation; morphological, gray level, and texture features were extracted and merged with a linear discriminant classifier to score the detected masses. For the DCNN-based CAD system, ROIs from five consecutive slices centered at each candidate were passed through the trained DCNN and a mass likelihood score was generated. The performances of the CAD systems were evaluated using free-response ROC curves and the performance difference was analyzed using a non-parametric method.RESULTSBefore transfer learning, the DCNN trained only on mammograms with an AUC of 0.99 classified DBT masses with an AUC of 0.81 in the DBT training set. After transfer learning with DBT, the AUC improved to 0.90. For breast-based CAD detection in the test set, the sensitivity for the feature-based and the DCNN-based CAD systems was 83% and 91%, respectively, at 1 FP/DBT volume. The difference between the performances for the two systems was statistically significant (p-value < 0.05).CONCLUSIONSThe image patterns learned from the mammograms were transferred to the mass detection on DBT slices through the DCNN. This study demonstrated that large data sets collected from mammography are useful for developing new CAD systems for DBT, alleviating the problem and effort of collecting entirely new large data sets for the new modality. |
| Author | Wei, Jun Cha, Kenny Samala, Ravi K. Helvie, Mark A. Hadjiiski, Lubomir Chan, Heang-Ping |
| Author_xml | – sequence: 1 givenname: Ravi K. surname: Samala fullname: Samala, Ravi K. email: rsamala@umich.edu organization: Department of Radiology, University of Michigan, Ann Arbor, Michigan 48109 – sequence: 2 givenname: Heang-Ping surname: Chan fullname: Chan, Heang-Ping organization: Department of Radiology, University of Michigan, Ann Arbor, Michigan 48109 – sequence: 3 givenname: Lubomir surname: Hadjiiski fullname: Hadjiiski, Lubomir organization: Department of Radiology, University of Michigan, Ann Arbor, Michigan 48109 – sequence: 4 givenname: Mark A. surname: Helvie fullname: Helvie, Mark A. organization: Department of Radiology, University of Michigan, Ann Arbor, Michigan 48109 – sequence: 5 givenname: Jun surname: Wei fullname: Wei, Jun organization: Department of Radiology, University of Michigan, Ann Arbor, Michigan 48109 – sequence: 6 givenname: Kenny surname: Cha fullname: Cha, Kenny organization: Department of Radiology, University of Michigan, Ann Arbor, Michigan 48109 |
| BackLink | https://www.ncbi.nlm.nih.gov/pubmed/27908154$$D View this record in MEDLINE/PubMed |
| BookMark | eNp9kc1u1DAUhS1URKeFBS-AvASktLZjxzYLJFTKj9QKFrC2PI49Y0jsYDszyts3w0yrgiiru7jfOffo3BNwFGKwADzH6AxjLM7xGZUNryl7BBaE8rqiBMkjsEBI0opQxI7BSc4_EEJNzdATcEy4RAIzugDTtc4ZtrZYU3wM0AfY-pUvuoPLZHUusMQ-5imUtc0-v4HvrR2giWETu3GnmMFgx_R7lG1MP-HWlzUsSYfsbIKd1Sn4sIIuxR72uu_jKulhPT0Fj53usn12mKfg-4fLbxefqqsvHz9fvLuqDBU1q9pWykZyLK3jRhhnMCHYkCVvNOcNMtIxJAyrlw3j2gna0Na0rhaNFVI4TOpT8HbvO4zL3rbGhjlbp4bke50mFbVXf26CX6tV3CiGa8Yxnw1eHgxS_DXaXFTvs7Fdp4ONY1ZYUCYIkWKHvrh_6-7Ibd8z8GoPmBRzTtbdIRip3S8VVodfzuz5X6yZ_7LrfI7pu38qqr1i6zs7PWytrr8e-Nd7Pt86_zfOg_Ampnvmw1z_DaNgzXM |
| CODEN | MPHYA6 |
| CitedBy_id | crossref_primary_10_1016_j_compbiomed_2025_110581 crossref_primary_10_1038_s41597_023_02025_1 crossref_primary_10_1016_j_imu_2021_100568 crossref_primary_10_1016_j_clinimag_2019_10_014 crossref_primary_10_32604_cmc_2021_018461 crossref_primary_10_1007_s00138_019_01042_8 crossref_primary_10_1109_ACCESS_2020_2972610 crossref_primary_10_1007_s10278_021_00530_6 crossref_primary_10_3389_fmolb_2020_599333 crossref_primary_10_1007_s13735_021_00218_1 crossref_primary_10_3390_jimaging8090231 crossref_primary_10_1016_j_media_2017_07_005 crossref_primary_10_1007_s12553_022_00693_4 crossref_primary_10_1016_j_ejmp_2019_07_017 crossref_primary_10_1016_j_compmedimag_2018_09_004 crossref_primary_10_1016_j_media_2018_03_006 crossref_primary_10_1038_s41598_017_15720_y crossref_primary_10_1088_1361_6528_adddc5 crossref_primary_10_1002_jmri_26534 crossref_primary_10_1007_s12194_019_00552_4 crossref_primary_10_1148_radiol_2018180547 crossref_primary_10_1007_s11277_022_09864_y crossref_primary_10_1016_j_crad_2018_01_001 crossref_primary_10_1109_ACCESS_2021_3056150 crossref_primary_10_1007_s00521_025_11215_4 crossref_primary_10_1002_mp_13550 crossref_primary_10_1088_1361_6560_aa93d4 crossref_primary_10_1007_s13246_020_00933_9 crossref_primary_10_1016_j_artmed_2018_04_005 crossref_primary_10_1016_j_ejmp_2021_03_021 crossref_primary_10_3390_app10228298 crossref_primary_10_1053_j_semnuclmed_2022_02_003 crossref_primary_10_3390_electronics12071583 crossref_primary_10_7717_peerj_cs_2076 crossref_primary_10_1109_TCBB_2020_2986544 crossref_primary_10_1155_2020_8860011 crossref_primary_10_1097_IJG_0000000000000988 crossref_primary_10_1109_ACCESS_2019_2891948 crossref_primary_10_1038_s41746_019_0172_3 crossref_primary_10_1002_mp_15883 crossref_primary_10_1016_j_acra_2017_12_025 crossref_primary_10_1016_j_cmpb_2023_107643 crossref_primary_10_1007_s12194_017_0406_5 crossref_primary_10_1007_s00330_019_06457_5 crossref_primary_10_1016_j_diii_2019_08_005 crossref_primary_10_1002_mp_14678 crossref_primary_10_1016_j_cmpb_2018_05_027 crossref_primary_10_1016_j_ibmed_2025_100281 crossref_primary_10_1186_s12859_019_2823_4 crossref_primary_10_1038_s41551_018_0305_z crossref_primary_10_1088_1361_6560_ab98d0 crossref_primary_10_1097_RCT_0000000000000928 crossref_primary_10_1016_j_ahj_2020_07_009 crossref_primary_10_1007_s11042_021_11775_2 crossref_primary_10_1088_1361_6560_ab5745 crossref_primary_10_1109_TMI_2018_2870343 crossref_primary_10_1145_3453166 crossref_primary_10_1016_j_ejmp_2018_12_012 crossref_primary_10_1016_j_media_2019_03_009 crossref_primary_10_1007_s11831_023_09899_9 crossref_primary_10_1016_j_neucom_2021_08_159 crossref_primary_10_3322_caac_21552 crossref_primary_10_1002_mp_13451 crossref_primary_10_1109_TMI_2023_3302799 crossref_primary_10_1002_mp_15631 crossref_primary_10_1155_2022_1367366 crossref_primary_10_1515_revneuro_2025_0062 crossref_primary_10_1016_j_compbiomed_2020_103698 crossref_primary_10_1088_1361_6560_aabb5b crossref_primary_10_1088_1361_6560_acaba7 crossref_primary_10_1088_2057_1976_ac7ada crossref_primary_10_1002_mco2_70247 crossref_primary_10_1148_rg_2019180124 crossref_primary_10_1002_cam4_70069 crossref_primary_10_1038_s41598_017_09315_w crossref_primary_10_1016_j_jbi_2020_103627 crossref_primary_10_1053_j_jvca_2024_12_016 crossref_primary_10_1148_ryai_220159 crossref_primary_10_3390_app11104573 crossref_primary_10_3390_computers12110220 crossref_primary_10_1002_jbio_201800255 crossref_primary_10_1117_1_JMI_11_2_024005 crossref_primary_10_1097_HCO_0000000000000833 crossref_primary_10_1109_ACCESS_2019_2892795 crossref_primary_10_1259_bjr_20170545 crossref_primary_10_1053_j_semnuclmed_2025_01_008 crossref_primary_10_1088_1361_6560_ab82e8 crossref_primary_10_1038_s41598_020_71914_x crossref_primary_10_1016_j_compbiomed_2022_105221 crossref_primary_10_1016_j_ins_2020_12_086 crossref_primary_10_1016_j_radi_2021_05_002 crossref_primary_10_1002_ett_4017 crossref_primary_10_3390_diagnostics13010045 crossref_primary_10_3389_fphar_2024_1437167 crossref_primary_10_1016_j_cmpb_2018_01_017 crossref_primary_10_1016_j_hlpt_2022_100702 crossref_primary_10_3233_THC_228017 crossref_primary_10_1002_mp_16188 crossref_primary_10_1080_21681163_2024_2302387 crossref_primary_10_1109_ACCESS_2020_3004056 crossref_primary_10_1001_jamanetworkopen_2021_19100 crossref_primary_10_1016_j_acra_2020_12_001 crossref_primary_10_1155_2022_9082694 crossref_primary_10_1007_s12530_019_09297_2 crossref_primary_10_1186_s12938_020_00831_x crossref_primary_10_1007_s13246_019_00822_w crossref_primary_10_1109_ACCESS_2021_3071297 crossref_primary_10_1016_j_compbiomed_2017_04_006 crossref_primary_10_1016_j_semcancer_2023_02_005 crossref_primary_10_3390_jcm12010385 crossref_primary_10_1007_s12559_020_09813_6 crossref_primary_10_1016_j_semcancer_2020_06_002 crossref_primary_10_1109_RBME_2024_3357877 crossref_primary_10_3390_diagnostics11040688 crossref_primary_10_1007_s11547_022_01512_6 crossref_primary_10_32604_cmes_2023_025484 crossref_primary_10_1016_j_bspc_2023_105263 crossref_primary_10_3389_fnins_2020_00858 crossref_primary_10_1016_j_neucom_2018_06_080 crossref_primary_10_1007_s41870_023_01308_5 crossref_primary_10_1002_mp_13264 crossref_primary_10_1016_j_bspc_2019_01_001 crossref_primary_10_1016_j_neucom_2018_06_084 crossref_primary_10_1016_j_acra_2022_12_038 crossref_primary_10_1016_j_bspc_2019_101566 crossref_primary_10_1259_bjr_20190580 crossref_primary_10_1088_2516_1091_acc4fc crossref_primary_10_1007_s11042_022_12599_4 crossref_primary_10_1016_j_compbiomed_2021_104873 crossref_primary_10_1088_1361_6560_ab553d crossref_primary_10_1148_radiol_2018171820 crossref_primary_10_1007_s00330_018_5978_x crossref_primary_10_1016_j_ejrad_2024_111457 crossref_primary_10_1002_aisy_202000063 crossref_primary_10_1016_j_jacr_2017_12_029 crossref_primary_10_1155_2020_7156165 crossref_primary_10_1016_j_jacr_2017_12_028 crossref_primary_10_3348_jksr_2020_0152 crossref_primary_10_1088_1555_6611_abb596 crossref_primary_10_1002_int_22622 crossref_primary_10_1109_ACCESS_2019_2939167 crossref_primary_10_1016_j_ymeth_2019_02_010 crossref_primary_10_1080_21681163_2023_2172615 crossref_primary_10_1002_ima_22423 crossref_primary_10_1109_ACCESS_2020_2984214 crossref_primary_10_1002_cac2_12215 crossref_primary_10_1088_1361_6560_aaef0a crossref_primary_10_1007_s11277_024_11428_1 crossref_primary_10_1111_all_15849 crossref_primary_10_1016_j_acra_2018_06_019 crossref_primary_10_1109_ACCESS_2020_3038909 crossref_primary_10_1016_j_bspc_2021_102976 crossref_primary_10_1371_journal_pone_0233166 crossref_primary_10_1016_j_compbiomed_2018_04_004 crossref_primary_10_1007_s11036_020_01672_7 crossref_primary_10_1007_s11042_023_16938_x crossref_primary_10_1016_j_compmedimag_2018_10_008 |
| Cites_doi | 10.1007/978-3-319-41546-8_7 10.1117/12.2217045 10.1118/1.1543934 10.1109/42.476112 10.1109/42.538937 10.1118/1.2968098 10.1117/12.2216289 10.1109/TMI.2016.2535302 10.1117/12.2208825 10.1148/radiol.2015141959 10.1148/radiol.2373041657 10.1117/12.710198 10.1007/978-3-540-70538-3_39 10.1109/ISBI.2007.356896 10.1118/1.597428 10.1109/34.31448 10.1118/1.2211710 10.1007/s11263‐015‐0816‐y 10.1117/12.2217260 10.1117/12.2217092 10.1016/j.acra.2007.02.017 10.1118/1.598805 10.1117/12.2043513 10.1016/j.acra.2009.11.011 10.1148/radiology.205.2.9356620 10.1109/CVPR.2009.5206848 10.1259/bjr/19385909 10.1016/0893‐6080(95)00061‐5 10.1056/NEJMsr1504363 10.1038/s41598-017-09315-w 10.1118/1.598228 10.1117/12.2006039 10.1148/radiol.11103514 10.1118/1.3432570 10.1088/0031‐9155/60/21/8457 10.1118/1.4791643 10.1258/jms.2012.012078 10.1109/TMI.2016.2528162 10.1117/12.2216852 10.1109/CVPR.2007.383157 10.1109/ISBI.2006.1625167 10.1088/0031‐9155/40/5/010 10.1088/0031‐9155/61/19/7092 10.1016/j.media.2015.08.001 10.1118/1.1800713 10.1117/12.2216327 10.1118/1.2237543 10.1118/1.4770279 10.1118/1.4944498 10.1080/21681163.2015.1124249 10.1118/1.1758791 10.1088/0031‐9155/59/23/7457 10.1118/1.3284974 10.1118/1.598389 10.1117/12.878214 |
| ContentType | Journal Article |
| Copyright | American Association of Physicists in Medicine 2016 American Association of Physicists in Medicine 2016 American Association of Physicists in Medicine. 2016 American Association of Physicists in Medicine |
| Copyright_xml | – notice: American Association of Physicists in Medicine – notice: 2016 American Association of Physicists in Medicine – notice: 2016 American Association of Physicists in Medicine. 2016 American Association of Physicists in Medicine |
| DBID | AAYXX CITATION CGR CUY CVF ECM EIF NPM 7X8 5PM |
| DOI | 10.1118/1.4967345 |
| DatabaseName | CrossRef Medline MEDLINE MEDLINE (Ovid) MEDLINE MEDLINE PubMed MEDLINE - Academic PubMed Central (Full Participant titles) |
| DatabaseTitle | CrossRef MEDLINE Medline Complete MEDLINE with Full Text PubMed MEDLINE (Ovid) MEDLINE - Academic |
| DatabaseTitleList | MEDLINE MEDLINE - Academic |
| Database_xml | – sequence: 1 dbid: NPM name: PubMed url: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed sourceTypes: Index Database – sequence: 2 dbid: 7X8 name: MEDLINE - Academic url: https://search.proquest.com/medline sourceTypes: Aggregation Database |
| DeliveryMethod | fulltext_linktorsrc |
| Discipline | Medicine Physics |
| DocumentTitleAlternate | DBT mass detection using deep convolutional neural network |
| EISSN | 2473-4209 |
| EndPage | 6666 |
| ExternalDocumentID | PMC5135717 27908154 10_1118_1_4967345 MP7345 |
| Genre | article Journal Article |
| GrantInformation_xml | – fundername: U.S. Public Health Service (USPHS) grantid: RO1 CA151443 funderid: http://dx.doi.org/10.13039/100007197 – fundername: U.S. Public Health Service (USPHS) funderid: RO1 CA151443 – fundername: NCI NIH HHS grantid: R01 CA151443 – fundername: ; ; grantid: RO1 CA151443 |
| GroupedDBID | --- --Z -DZ .GJ 0R~ 1OB 1OC 29M 2WC 33P 36B 3O- 4.4 476 53G 5GY 5RE 5VS AAHHS AANLZ AAQQT AASGY AAXRX AAZKR ABCUV ABEFU ABFTF ABJNI ABLJU ABQWH ABTAH ABXGK ACAHQ ACBEA ACCFJ ACCZN ACGFO ACGFS ACGOF ACPOU ACSMX ACXBN ACXQS ADBBV ADBTR ADKYN ADOZA ADXAS ADZMN AEEZP AEGXH AEIGN AENEX AEQDE AEUYR AFBPY AFFPM AHBTC AIACR AIAGR AIURR AIWBW AJBDE ALMA_UNASSIGNED_HOLDINGS ALUQN AMYDB ASPBG BFHJK C45 CS3 DCZOG DRFUL DRMAN DRSTM DU5 EBD EBS EJD EMB EMOBN F5P G8K HDBZQ HGLYW I-F KBYEO LATKE LEEKS LOXES LUTES LYRES MEWTI O9- OVD P2P P2W PALCI PHY RJQFR RNS ROL SAMSI SUPJJ SV3 TEORI TN5 TWZ USG WOHZO WXSBR XJT ZGI ZVN ZXP ZY4 ZZTAW AAHQN AAIPD AAMNL AAYCA ABDPE AFWVQ AITYG ALVPJ AAMMB AAYXX ABUFD ADMLS AEFGJ AEYWJ AGHNM AGXDD AGYGG AIDQK AIDYY AIQQE CITATION LH4 CGR CUY CVF ECM EIF NPM 7X8 5PM |
| ID | FETCH-LOGICAL-c4835-dd9969719ef7c8cfc1221c2b76a7760c9f508c53b657af8464dcdf386e898f123 |
| IEDL.DBID | DRFUL |
| ISICitedReferencesCount | 196 |
| ISICitedReferencesURI | http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000390237200043&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D |
| ISSN | 0094-2405 |
| IngestDate | Tue Nov 04 01:54:43 EST 2025 Thu Oct 02 11:46:28 EDT 2025 Mon Jul 21 06:06:04 EDT 2025 Tue Nov 18 20:59:43 EST 2025 Sat Nov 29 07:53:35 EST 2025 Wed Jan 22 16:30:48 EST 2025 Fri Jun 21 00:14:38 EDT 2024 Sun Jul 14 10:28:25 EDT 2019 |
| IsDoiOpenAccess | false |
| IsOpenAccess | true |
| IsPeerReviewed | true |
| IsScholarly | true |
| Issue | 12 |
| Keywords | digital breast tomosynthesis transfer learning deep-learning computer-aided detection convolutional neural network mass |
| Language | English |
| License | 0094-2405/2016/43(12)/6654/13/$30.00 http://onlinelibrary.wiley.com/termsAndConditions#vor |
| LinkModel | DirectLink |
| MergedId | FETCHMERGED-LOGICAL-c4835-dd9969719ef7c8cfc1221c2b76a7760c9f508c53b657af8464dcdf386e898f123 |
| Notes | Telephone: (734) 647‐8556; Fax: (734) 615‐5513. Author to whom correspondence should be addressed. Electronic mail rsamala@umich.edu ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 23 Author to whom correspondence should be addressed. Electronic mail: rsamala@umich.edu; Telephone: (734) 647-8556; Fax: (734) 615-5513. |
| ORCID | 0000-0002-6661-4801 |
| OpenAccessLink | http://doi.org/10.1118/1.4967345 |
| PMID | 27908154 |
| PQID | 1845822987 |
| PQPubID | 23479 |
| PageCount | 13 |
| ParticipantIDs | pubmed_primary_27908154 scitation_primary_10_1118_1_4967345 wiley_primary_10_1118_1_4967345_MP7345 pubmedcentral_primary_oai_pubmedcentral_nih_gov_5135717 crossref_primary_10_1118_1_4967345 crossref_citationtrail_10_1118_1_4967345 proquest_miscellaneous_1845822987 |
| PublicationCentury | 2000 |
| PublicationDate | December 2016 |
| PublicationDateYYYYMMDD | 2016-12-01 |
| PublicationDate_xml | – month: 12 year: 2016 text: December 2016 |
| PublicationDecade | 2010 |
| PublicationPlace | United States |
| PublicationPlace_xml | – name: United States |
| PublicationTitle | Medical physics (Lancaster) |
| PublicationTitleAlternate | Med Phys |
| PublicationYear | 2016 |
| Publisher | American Association of Physicists in Medicine |
| Publisher_xml | – name: American Association of Physicists in Medicine |
| References | Ciompi, de Hoop, van Riel, Chung, Scholten, Oudkerk, de Jong, Prokop, van Ginneken (c26) 2015 Wu, Zhou, Hadjiiski, Shi, Wei, Paramagul, Sahiner, Chan (c39) 2007 Samala, Chan, Lu, Hadjiiski, Wei, Helvie (c19) 2015 Ge, Sahiner, Hadjiiski, Chan, Wei, Helvie, Zhou (c17) 2006 Niklason (c6) 1997 Fukunaga, Hayes (c9) 1989 Burgess (c54) 2004 Bernardi, Ciatto, Pellegrini, Anesi, Burlon, Cauli, Depaoli, Larentis, Malesani, Targa (c29) 2014 Zuley, Bandos, Abrams, Cohen, Hakim, Sumkin, Drescher, Rockette, Gur (c27) 2010 Gao, Bagci, Lu, Wu, Buty, Shin, Roth, Papadakis, Depeursinge, Summers (c25) 2016 Lauby-Secretan, Scoccianti, Loomis, Benbrahim-Tallaa, Bouvard, Bianchini, Straif (c5) 2015 Way, Sahiner, Hadjiiski, Chan (c42) 2010 Chan, Sahiner, Lam, Petrick, Helvie, Goodsitt, Adler (c43) 1998 Wei (c37) 2007 Broeders, Moss, Nyström, Njor, Jonsson, Paap, Massat, Duffy, Lynge, Paci (c4) 2012 Chan, Lo, Sahiner, Lam, Helvie (c15) 1995 Chan, Wei, Zhang, Helvie, Moore, Sahiner, Hadjiiski, Kopans (c32) 2008 Sechopoulos (c8) 2013 Wallis, Moa, Zanca, Leifland, Danielsson (c30) 2012 Cha, Hadjiiski, Samala, Chan, Cohan, Caoili (c48) 2016 van Schie, Wallis, Leifland, Danielsson, Karssemeijer (c62) 2013 Chan, Wei, Sahiner, Rafferty, Wu, Roubidoux, Moore, Kopans, Hadjiiski, Helvie (c31) 2005 Astley (c28) 2013 Wu (c7) 2003 Shin, Roth, Gao, Lu, Xu, Nogues, Yao, Mollura, Summers (c11) 2016 Russakovsky, Deng, Su, Krause, Satheesh, Ma, Huang, Karpathy, Khosla, Bernstein (c1) 2015 Tajbakhsh, Shin, Gurudu, Hurst, Kendall, Gotway, Liang (c14) 2016 Lo, Lou, Lin, Freedman, Mun (c20) 1995 Chan, Wu, Sahiner, Wei, Helvie, Zhang, Moore, Kopans, Hadjiiski, Way (c35) 2010 Srivastava, Hinton, Krizhevsky, Sutskever, Salakhutdinov (c53) 2014 Zhou, Chan, Paramagul, Roubidoux, Sahiner, Hadjiiski, Petrick (c38) 2004 Näppi, Hironaka, Regge, Yoshida (c12) 2016 Sahiner, Chan, Petrick, Wei, Helvie, Adler, Goodsitt (c16) 1996 Lo, Chan, Lin, Li, Freedman, Mun (c21) 1995 Samala, Wei, Chan, Hadjiiski, Cha, Helvie (c40) 2016 Sahiner, Chan, Petrick, Helvie, Goodsitt (c41) 1998 Samala, Chan, Lu, Hadjiiski, Wei, Helvie (c18) 2014 Cha, Hadjiiski, Samala, Chan, Caoili, Cohan (c24) 2016 Cha, Hadjiiski, Samala, Chan, Cohan, Caoili, Paramagul, Alva, Weizer (c50) 2016 Fotin, Yin, Haldankar, Hoffmeister, Periaswamy (c60) 2016 Cruz-Roa, Arévalo, Judkins, Madabhushi, González (c13) 2015 Zhang, Chan, Sahiner, Wei, Goodsitt, Hadjiiski, Ge, Zhou (c36) 2006 Samala, Chan, Hadjiiski, Helvie (c22) 2016 Samala, Chan, Hadjiiski, Cha, Helvie (c52) 2016 Chan, Sahiner, Wagner, Petrick (c10) 1999 Morra, Sacchetto, Durando, Agliozzo, Carbonaro, Delsanto, Pesce, Persano, Mariscotti, Marra (c61) 2015 Cha, Hadjiiski, Chan, Caoili, Cohan, Weizer, Zhou (c51) 2016 Krizhevsky, Sutskever, Hinton (c45) 2012 Wei, Chan, Sahiner, Hadjiiski, Helvie, Zhou, Lu (c33) 2011 Chan, Wei, Helvie, Sahiner, Adler, Goodsitt, Petrick (c55) 1995 Samala, Chan, Lu, Hadjiiski, Wei, Helvie (c64) 2014 2015; 9681 2012; 262 2010; 37 2012 2014; 9035 2006; 33 1995; 14 2010; 17 2013; 40 2007; 6512 1999; 26 2005; 237 2013; 8673 2009 2008 2007 2012; 19 2008; 35 2006 2014; 85 2003; 30 1996; 15 2016; 35 1995; 8 1998; 25 2015; 372 2011; 7963 2016; 9785 2015; 26 1995; 40 2004; 31 1997; 205 1989; 11 2015; 115 2015; 60 2001 2015; 277 1995; 22 2016; 43 2014; 59 2014; 15 2007; 6 2016; 61 2016 2015 2014 e_1_2_8_28_1 e_1_2_8_24_1 e_1_2_8_47_1 e_1_2_8_26_1 e_1_2_8_49_1 Krizhevsky A. (e_1_2_8_46_1) 2012 Srivastava N. (e_1_2_8_54_1) 2014; 15 e_1_2_8_5_1 e_1_2_8_7_1 e_1_2_8_9_1 e_1_2_8_20_1 e_1_2_8_43_1 e_1_2_8_66_1 e_1_2_8_22_1 e_1_2_8_45_1 e_1_2_8_62_1 e_1_2_8_41_1 e_1_2_8_60_1 e_1_2_8_17_1 e_1_2_8_19_1 Chan H.‐P. (e_1_2_8_64_1) 2008 e_1_2_8_13_1 e_1_2_8_36_1 e_1_2_8_59_1 e_1_2_8_15_1 e_1_2_8_38_1 e_1_2_8_57_1 e_1_2_8_32_1 e_1_2_8_55_1 e_1_2_8_11_1 e_1_2_8_34_1 e_1_2_8_53_1 e_1_2_8_30_1 e_1_2_8_29_1 e_1_2_8_25_1 e_1_2_8_27_1 e_1_2_8_48_1 Heath M. (e_1_2_8_35_1) 2001 e_1_2_8_2_1 e_1_2_8_4_1 e_1_2_8_6_1 Yosinski J. (e_1_2_8_3_1) 2014 e_1_2_8_8_1 e_1_2_8_21_1 e_1_2_8_42_1 e_1_2_8_23_1 e_1_2_8_44_1 e_1_2_8_65_1 e_1_2_8_63_1 e_1_2_8_40_1 e_1_2_8_61_1 e_1_2_8_18_1 e_1_2_8_39_1 e_1_2_8_14_1 e_1_2_8_16_1 e_1_2_8_37_1 e_1_2_8_58_1 Cha K. H. (e_1_2_8_51_1) 2016 e_1_2_8_10_1 e_1_2_8_31_1 e_1_2_8_56_1 e_1_2_8_12_1 e_1_2_8_33_1 e_1_2_8_52_1 e_1_2_8_50_1 |
| References_xml | – start-page: 1285 year: 2016 ident: c11 article-title: Deep convolutional neural networks for computer-aided detection: CNN architectures, dataset characteristics and transfer learning publication-title: IEEE Trans. Med. Imaging – start-page: 2871 year: 2004 ident: c38 article-title: Computerized nipple identification for multiple image analysis in computer-aided diagnosis publication-title: Med. Phys. – start-page: 1987 year: 2004 ident: c54 article-title: On the noise variance of a digital mammography system publication-title: Med. Phys. – start-page: 1201 year: 1995 ident: c21 article-title: Artificial convolution neural network for medical image pattern recognition publication-title: Neural Networks – start-page: 97850X year: 2016 ident: c60 article-title: Detection of soft tissue densities from digital breast tomosynthesis: Comparison of conventional and deep learning approaches publication-title: Proc. SPIE – start-page: 56 year: 2015 ident: c61 article-title: Breast cancer: Computer-aided detection with digital breast tomosynthesis publication-title: Radiology – start-page: 788 year: 2012 ident: c30 article-title: Two-view and single-view tomosynthesis versus full-field digital mammography: High-resolution x-ray imaging observer study publication-title: Radiology – start-page: 796311-1 year: 2011 ident: c33 article-title: Computer-aided detection of breast masses in digital breast tomosynthesis (DBT): Improvement of false positive reduction by optimization of object segmentation publication-title: Proc. SPIE – start-page: 1097 year: 2012 ident: c45 article-title: Imagenet classification with deep convolutional neural networks publication-title: Adv. Neural Inf. Proc. Syst. – start-page: 1555 year: 1995 ident: c15 article-title: Computer-aided detection of mammographic microcalcifications: Pattern recognition with an artificial neural network publication-title: Med. Phys. – start-page: 873 year: 1989 ident: c9 article-title: Effects of sample size on classifier design publication-title: IEEE Trans. Pattern Anal. Mach. Intell. – start-page: 65122U1 year: 2007 ident: c39 article-title: A dynamic multiple thresholding method for automated breast boundary detection in digitized mammograms publication-title: Proc. SPIE – start-page: 14 year: 2012 ident: c4 article-title: The impact of mammographic screening on breast cancer mortality in Europe: A review of observational studies publication-title: J. Med. Screening – start-page: e1174 year: 2014 ident: c29 article-title: Application of breast tomosynthesis in screening: Incremental effect on mammography acquisition and reading time publication-title: Br. J. Radiol. – start-page: 7457 year: 2014 ident: c18 article-title: Digital breast tomosynthesis: Computer-aided detection of clustered microcalcifications on planar projection images publication-title: Phys. Med. Biol. – start-page: 2975 year: 2006 ident: c17 article-title: Computer aided detection of clusters of microcalcifications on full field digital mammograms publication-title: Med. Phys. – start-page: 97850Y-1 year: 2016 ident: c52 article-title: Deep-learning convolution neural network for computer-aided detection of microcalcifications in digital breast tomosynthesis publication-title: Proc. SPIE – start-page: 399 year: 1997 ident: c6 article-title: Digital tomosynthesis in breast imaging publication-title: Radiology – start-page: 3781 year: 2006 ident: c36 article-title: A comparative study of limited-angle cone-beam reconstruction methods for breast tomosynthesis publication-title: Med. Phys. – start-page: 978512-1 year: 2016 ident: c48 article-title: Comparison of bladder segmentation using deep-learning convolutional neural network with and without level sets publication-title: Proc. SPIE – start-page: 907 year: 2010 ident: c42 article-title: Effect of finite sample size on feature selection and classification: A simulation study publication-title: Med. Phys. – year: 2016 ident: c50 article-title: Bladder cancer segmentation in CT for treatment response assessment: Application of deep-learning convolution neural network—A pilot study publication-title: Tomography – start-page: 014301 year: 2013 ident: c8 article-title: A review of breast tomosynthesis. Part I. The image acquisition process publication-title: Med. Phys. – start-page: 211 year: 2015 ident: c1 article-title: ImageNet large scale visual recognition challenge publication-title: Int. J. Comput. Vision – start-page: 659 year: 2007 ident: c37 article-title: Computer aided detection systems for breast masses: Comparison of performances on full-field digital mammograms and digitized screen-film mammograms publication-title: Acad. Radiol. – start-page: 8457 year: 2015 ident: c19 article-title: Computer-aided detection system for clustered microcalcifications in digital breast tomosynthesis using joint information from volumetric and planar projection images publication-title: Phys. Med. Biol. – start-page: 3576 year: 2010 ident: c35 article-title: Characterization of masses in digital breast tomosynthesis: Comparison of machine learning in projection views and reconstructed slices publication-title: Med. Phys. – start-page: 2007 year: 1998 ident: c43 article-title: Computerized analysis of mammographic microcalcifications in morphological and texture feature space publication-title: Med. Phys. – start-page: 2353 year: 2015 ident: c5 article-title: Breast-cancer screening—Viewpoint of the IARC Working Group publication-title: N. Engl. J. Med. – start-page: 4087 year: 2008 ident: c32 article-title: Computer-aided detection of masses in digital tomosynthesis mammography: Comparison of three approaches publication-title: Med. Phys. – start-page: 365 year: 2003 ident: c7 article-title: Tomographic mammography using a limited number of low-dose cone-beam projection images publication-title: Med. Phys. – start-page: 978523-1 year: 2016 ident: c40 article-title: First and second-order features for detection of masses in digital breast tomosynthesis publication-title: Proc. SPIE – start-page: 1929 year: 2014 ident: c53 article-title: Dropout: A simple way to prevent neural networks from overfitting publication-title: J. Mach. Learn. Res. – start-page: 711 year: 1995 ident: c20 article-title: Artificial convolution neural network techniques and applications to lung nodule detection publication-title: IEEE Trans. Med. Imaging – start-page: 1 year: 2016 ident: c25 article-title: Holistic classification of CT attenuation patterns for interstitial lung diseases via deep convolutional neural networks publication-title: Comput. Methods Biomech. Biomed. Eng.: Imaging Visualization – start-page: 1299 year: 2016 ident: c14 article-title: Convolutional neural networks for medical image Analysis: Full training or fine tuning? publication-title: IEEE Trans. Med. Imaging – start-page: 598 year: 1996 ident: c16 article-title: Classification of mass and normal breast tissue: A convolution neural network classifier with spatial domain and texture images publication-title: IEEE Trans. Med. Imaging – start-page: 516 year: 1998 ident: c41 article-title: Computerized characterization of masses on mammograms: The rubber band straightening transform and texture analysis publication-title: Med. Phys. – start-page: 195 year: 2015 ident: c26 article-title: Automatic classification of pulmonary peri-fissural nodules in computed tomography using an ensemble of 2D views and a convolutional neural network out-of-the-box publication-title: Med. Image Anal. – start-page: 97853W-1 year: 2016 ident: c51 article-title: Computer-aided detection of bladder mass within non-contrast-enhanced region of CT Urography (CTU) publication-title: Proc. SPIE – start-page: 7092 year: 2016 ident: c22 article-title: Analysis of computer-aided detection techniques and signal characteristics for clustered microcalcifications on digital mammography and digital breast tomosynthesis publication-title: Phys. Med. Biol. – start-page: 1075 year: 2005 ident: c31 article-title: Computer-aided detection system for breast masses on digital tomosynthesis mammograms–preliminary experience publication-title: Radiology – start-page: 2654 year: 1999 ident: c10 article-title: Classifier design for computer-aided diagnosis: Effects of finite sample size on the mean performance of classical and neural network classifiers publication-title: Med. Phys. – start-page: 857 year: 1995 ident: c55 article-title: Computer-aided classification of mammographic masses and normal tissue: Linear discriminant analysis in texture feature space publication-title: Phys. Med. Biol. – start-page: 97852B-1 year: 2016 ident: c12 article-title: Deep transfer learning of virtual endoluminal views for the detection of polyps in CT colonography publication-title: Proc. SPIE – start-page: 041902 year: 2013 ident: c62 article-title: Mass detection in reconstructed digital breast tomosynthesis volumes with a computer-aided detection system trained on 2D mammograms publication-title: Med. Phys. – start-page: 1882 year: 2016 ident: c24 article-title: Urinary bladder segmentation in CT urography using deep-learning convolutional neural network and level sets publication-title: Med. Phys. – start-page: 450 year: 2010 ident: c27 article-title: Time to diagnosis and performance levels during repeat interpretations of digital breast tomosynthesis: Preliminary observations publication-title: Acad. Radiol. – start-page: 968103-1 year: 2015 ident: c13 article-title: A method for medulloblastoma tumor differentiation based on convolutional neural networks and transfer learning publication-title: Proc. SPIE – start-page: 86730S-1 year: 2013 ident: c28 article-title: A comparison of image interpretation times in full field digital mammography and digital breast tomosynthesis publication-title: Proc. SPIE – start-page: 90350Y year: 2014 ident: c64 article-title: Digital breast tomosynthesis: Effects of projection-view distribution on computer-aided detection of microcalcification clusters publication-title: Proc. SPIE – volume: 35 start-page: 4087 year: 2008 end-page: 4095 article-title: Computer‐aided detection of masses in digital tomosynthesis mammography: Comparison of three approaches publication-title: Med. Phys. – volume: 37 start-page: 3576 year: 2010 end-page: 3586 article-title: Characterization of masses in digital breast tomosynthesis: Comparison of machine learning in projection views and reconstructed slices publication-title: Med. Phys. – start-page: 1097 year: 2012 end-page: 1105 article-title: Imagenet classification with deep convolutional neural networks publication-title: Adv. Neural Inf. Proc. Syst. – start-page: 248 year: 2009 end-page: 255 article-title: Imagenet: A large‐scale hierarchical image database – volume: 237 start-page: 1075 year: 2005 end-page: 1080 article-title: Computer‐aided detection system for breast masses on digital tomosynthesis mammograms–preliminary experience publication-title: Radiology – volume: 30 start-page: 365 year: 2003 end-page: 380 article-title: Tomographic mammography using a limited number of low‐dose cone‐beam projection images publication-title: Med. Phys. – volume: 31 start-page: 2871 year: 2004 end-page: 2882 article-title: Computerized nipple identification for multiple image analysis in computer‐aided diagnosis publication-title: Med. Phys. – volume: 40 start-page: 857 year: 1995 end-page: 876 article-title: Computer‐aided classification of mammographic masses and normal tissue: Linear discriminant analysis in texture feature space publication-title: Phys. Med. Biol. – volume: 9785 start-page: 97853W‐1 year: 2016 end-page: 97853W‐6 article-title: Computer‐aided detection of bladder mass within non‐contrast‐enhanced region of CT Urography (CTU) publication-title: Proc. SPIE – volume: 25 start-page: 516 year: 1998 end-page: 526 article-title: Computerized characterization of masses on mammograms: The rubber band straightening transform and texture analysis publication-title: Med. Phys. – year: 2016 article-title: Bladder cancer segmentation in CT for treatment response assessment: Application of deep‐learning convolution neural network—A pilot study publication-title: Tomography – volume: 43 start-page: 1882 year: 2016 end-page: 1896 article-title: Urinary bladder segmentation in CT urography using deep‐learning convolutional neural network and level sets publication-title: Med. Phys. – start-page: 51 year: 2016 end-page: 56 article-title: A comparison between a deep convolutional neural network and radiologists for classifying regions of interest in mammography – volume: 9785 start-page: 978512‐1 year: 2016 end-page: 978512‐7 article-title: Comparison of bladder segmentation using deep‐learning convolutional neural network with and without level sets publication-title: Proc. SPIE – volume: 60 start-page: 8457 year: 2015 end-page: 8479 article-title: Computer‐aided detection system for clustered microcalcifications in digital breast tomosynthesis using joint information from volumetric and planar projection images publication-title: Phys. Med. Biol. – volume: 115 start-page: 211 year: 2015 end-page: 252 article-title: ImageNet large scale visual recognition challenge publication-title: Int. J. Comput. Vision – start-page: 492 year: 2007 end-page: 495 article-title: Advantages and examples of resampling for CAD evaluation – volume: 17 start-page: 450 year: 2010 end-page: 455 article-title: Time to diagnosis and performance levels during repeat interpretations of digital breast tomosynthesis: Preliminary observations publication-title: Acad. Radiol. – volume: 22 start-page: 1555 year: 1995 end-page: 1567 article-title: Computer‐aided detection of mammographic microcalcifications: Pattern recognition with an artificial neural network publication-title: Med. Phys. – volume: 37 start-page: 907 year: 2010 end-page: 920 article-title: Effect of finite sample size on feature selection and classification: A simulation study publication-title: Med. Phys. – volume: 59 start-page: 7457 year: 2014 end-page: 7477 article-title: Digital breast tomosynthesis: Computer‐aided detection of clustered microcalcifications on planar projection images publication-title: Phys. Med. Biol. – volume: 15 start-page: 1929 year: 2014 end-page: 1958 article-title: Dropout: A simple way to prevent neural networks from overfitting publication-title: J. Mach. Learn. Res. – volume: 11 start-page: 873 year: 1989 end-page: 885 article-title: Effects of sample size on classifier design publication-title: IEEE Trans. Pattern Anal. Mach. Intell. – volume: 9785 start-page: 97850Y‐1 year: 2016 end-page: 97850Y‐7 article-title: Deep‐learning convolution neural network for computer‐aided detection of microcalcifications in digital breast tomosynthesis publication-title: Proc. SPIE – volume: 372 start-page: 2353 year: 2015 end-page: 2358 article-title: Breast‐cancer screening—Viewpoint of the IARC Working Group publication-title: N. Engl. J. Med. – volume: 40 start-page: 041902 year: 2013 article-title: Mass detection in reconstructed digital breast tomosynthesis volumes with a computer‐aided detection system trained on 2D mammograms publication-title: Med. Phys. – volume: 26 start-page: 195 year: 2015 end-page: 202 article-title: Automatic classification of pulmonary peri‐fissural nodules in computed tomography using an ensemble of 2D views and a convolutional neural network out‐of‐the‐box publication-title: Med. Image Anal. – volume: 85 start-page: e1174 year: 2014 end-page: e1178 article-title: Application of breast tomosynthesis in screening: Incremental effect on mammography acquisition and reading time publication-title: Br. J. Radiol. – start-page: 279 year: 2008 end-page: 285 article-title: Detection of masses in digital breast tomosynthesis mammography: Effects of the number of projection views and dose – volume: 9785 start-page: 97850X year: 2016 end-page: 97850X‐6 article-title: Detection of soft tissue densities from digital breast tomosynthesis: Comparison of conventional and deep learning approaches publication-title: Proc. SPIE – volume: 262 start-page: 788 year: 2012 end-page: 796 article-title: Two‐view and single‐view tomosynthesis versus full‐field digital mammography: High‐resolution x‐ray imaging observer study publication-title: Radiology – year: 2001 article-title: The digital database for screening mammography – year: 2015 – start-page: 1 year: 2007 end-page: 8 article-title: Unsupervised learning of invariant feature hierarchies with applications to object recognition – volume: 61 start-page: 7092 year: 2016 end-page: 7112 article-title: Analysis of computer‐aided detection techniques and signal characteristics for clustered microcalcifications on digital mammography and digital breast tomosynthesis publication-title: Phys. Med. Biol. – volume: 35 start-page: 1285 year: 2016 end-page: 1298 article-title: Deep convolutional neural networks for computer‐aided detection: CNN architectures, dataset characteristics and transfer learning publication-title: IEEE Trans. Med. Imaging – volume: 6512 start-page: 65122U1 year: 2007 end-page: 65122U8 article-title: A dynamic multiple thresholding method for automated breast boundary detection in digitized mammograms publication-title: Proc. SPIE – volume: 40 start-page: 014301 year: 2013 article-title: A review of breast tomosynthesis. Part I. The image acquisition process publication-title: Med. Phys. – volume: 9035 start-page: 90350Y year: 2014 article-title: Digital breast tomosynthesis: Effects of projection‐view distribution on computer‐aided detection of microcalcification clusters publication-title: Proc. SPIE – start-page: 560 year: 2008 article-title: Digital breast tomosynthesis (DBT) mammography: Effect of number of projection views on computerized mass detection using 2D and 3D approaches – volume: 9681 start-page: 968103‐1 year: 2015 end-page: 968103‐8 article-title: A method for medulloblastoma tumor differentiation based on convolutional neural networks and transfer learning publication-title: Proc. SPIE – volume: 9785 start-page: 978523‐1 year: 2016 end-page: 978523‐7 article-title: First and second‐order features for detection of masses in digital breast tomosynthesis publication-title: Proc. SPIE – year: 2016 – start-page: 1312 year: 2006 end-page: 1315 article-title: Comparing image detection algorithms using resampling – volume: 35 start-page: 1299 year: 2016 end-page: 1312 article-title: Convolutional neural networks for medical image Analysis: Full training or fine tuning? publication-title: IEEE Trans. Med. Imaging – volume: 205 start-page: 399 year: 1997 end-page: 406 article-title: Digital tomosynthesis in breast imaging publication-title: Radiology – year: 2012 – volume: 9785 start-page: 97852B‐1 year: 2016 end-page: 97852B‐8 article-title: Deep transfer learning of virtual endoluminal views for the detection of polyps in CT colonography publication-title: Proc. SPIE – volume: 15 start-page: 598 year: 1996 end-page: 610 article-title: Classification of mass and normal breast tissue: A convolution neural network classifier with spatial domain and texture images publication-title: IEEE Trans. Med. Imaging – volume: 33 start-page: 2975 year: 2006 end-page: 2988 article-title: Computer aided detection of clusters of microcalcifications on full field digital mammograms publication-title: Med. Phys. – volume: 25 start-page: 2007 year: 1998 end-page: 2019 article-title: Computerized analysis of mammographic microcalcifications in morphological and texture feature space publication-title: Med. Phys. – volume: 277 start-page: 56 year: 2015 end-page: 63 article-title: Breast cancer: Computer‐aided detection with digital breast tomosynthesis publication-title: Radiology – volume: 31 start-page: 1987 year: 2004 end-page: 1995 article-title: On the noise variance of a digital mammography system publication-title: Med. Phys. – volume: 6 start-page: 659 year: 2007 end-page: 669 article-title: Computer aided detection systems for breast masses: Comparison of performances on full‐field digital mammograms and digitized screen‐film mammograms publication-title: Acad. Radiol. – volume: 14 start-page: 711 year: 1995 end-page: 718 article-title: Artificial convolution neural network techniques and applications to lung nodule detection publication-title: IEEE Trans. Med. Imaging – volume: 26 start-page: 2654 year: 1999 end-page: 2668 article-title: Classifier design for computer‐aided diagnosis: Effects of finite sample size on the mean performance of classical and neural network classifiers publication-title: Med. Phys. – volume: 7963 start-page: 796311‐1 year: 2011 end-page: 796311‐6 article-title: Computer‐aided detection of breast masses in digital breast tomosynthesis (DBT): Improvement of false positive reduction by optimization of object segmentation publication-title: Proc. SPIE – start-page: 1 year: 2016 end-page: 6 article-title: Holistic classification of CT attenuation patterns for interstitial lung diseases via deep convolutional neural networks publication-title: Comput. Methods Biomech. Biomed. Eng.: Imaging Visualization – volume: 33 start-page: 3781 year: 2006 end-page: 3795 article-title: A comparative study of limited‐angle cone‐beam reconstruction methods for breast tomosynthesis publication-title: Med. Phys. – volume: 8673 start-page: 86730S‐1 year: 2013 end-page: 86730S‐8 article-title: A comparison of image interpretation times in full field digital mammography and digital breast tomosynthesis publication-title: Proc. SPIE – start-page: 3320 year: 2014 end-page: 3328 article-title: How transferable are features in deep neural networks? – volume: 19 start-page: 14 year: 2012 end-page: 25 article-title: The impact of mammographic screening on breast cancer mortality in Europe: A review of observational studies publication-title: J. Med. Screening – volume: 8 start-page: 1201 year: 1995 end-page: 1214 article-title: Artificial convolution neural network for medical image pattern recognition publication-title: Neural Networks – ident: e_1_2_8_24_1 doi: 10.1007/978-3-319-41546-8_7 – ident: e_1_2_8_61_1 doi: 10.1117/12.2217045 – ident: e_1_2_8_8_1 doi: 10.1118/1.1543934 – ident: e_1_2_8_59_1 – start-page: 560 volume-title: RSNA Program Book, 2008 year: 2008 ident: e_1_2_8_64_1 – ident: e_1_2_8_21_1 doi: 10.1109/42.476112 – ident: e_1_2_8_17_1 doi: 10.1109/42.538937 – ident: e_1_2_8_33_1 doi: 10.1118/1.2968098 – ident: e_1_2_8_52_1 doi: 10.1117/12.2216289 – ident: e_1_2_8_15_1 doi: 10.1109/TMI.2016.2535302 – ident: e_1_2_8_14_1 doi: 10.1117/12.2208825 – ident: e_1_2_8_62_1 doi: 10.1148/radiol.2015141959 – ident: e_1_2_8_32_1 doi: 10.1148/radiol.2373041657 – ident: e_1_2_8_4_1 – ident: e_1_2_8_40_1 doi: 10.1117/12.710198 – start-page: 1097 year: 2012 ident: e_1_2_8_46_1 article-title: Imagenet classification with deep convolutional neural networks publication-title: Adv. Neural Inf. Proc. Syst. – ident: e_1_2_8_66_1 doi: 10.1007/978-3-540-70538-3_39 – ident: e_1_2_8_57_1 doi: 10.1109/ISBI.2007.356896 – ident: e_1_2_8_16_1 doi: 10.1118/1.597428 – ident: e_1_2_8_10_1 doi: 10.1109/34.31448 – ident: e_1_2_8_18_1 doi: 10.1118/1.2211710 – ident: e_1_2_8_2_1 doi: 10.1007/s11263‐015‐0816‐y – ident: e_1_2_8_13_1 doi: 10.1117/12.2217260 – ident: e_1_2_8_53_1 doi: 10.1117/12.2217092 – ident: e_1_2_8_38_1 doi: 10.1016/j.acra.2007.02.017 – ident: e_1_2_8_11_1 doi: 10.1118/1.598805 – year: 2016 ident: e_1_2_8_51_1 article-title: Bladder cancer segmentation in CT for treatment response assessment: Application of deep‐learning convolution neural network—A pilot study publication-title: Tomography – ident: e_1_2_8_65_1 doi: 10.1117/12.2043513 – ident: e_1_2_8_28_1 doi: 10.1016/j.acra.2009.11.011 – ident: e_1_2_8_7_1 doi: 10.1148/radiology.205.2.9356620 – ident: e_1_2_8_47_1 doi: 10.1109/CVPR.2009.5206848 – ident: e_1_2_8_30_1 doi: 10.1259/bjr/19385909 – ident: e_1_2_8_22_1 doi: 10.1016/0893‐6080(95)00061‐5 – ident: e_1_2_8_6_1 doi: 10.1056/NEJMsr1504363 – ident: e_1_2_8_48_1 – ident: e_1_2_8_50_1 doi: 10.1038/s41598-017-09315-w – ident: e_1_2_8_42_1 doi: 10.1118/1.598228 – ident: e_1_2_8_29_1 doi: 10.1117/12.2006039 – ident: e_1_2_8_31_1 doi: 10.1148/radiol.11103514 – ident: e_1_2_8_36_1 doi: 10.1118/1.3432570 – ident: e_1_2_8_20_1 doi: 10.1088/0031‐9155/60/21/8457 – ident: e_1_2_8_63_1 doi: 10.1118/1.4791643 – ident: e_1_2_8_5_1 doi: 10.1258/jms.2012.012078 – ident: e_1_2_8_12_1 doi: 10.1109/TMI.2016.2528162 – ident: e_1_2_8_49_1 doi: 10.1117/12.2216852 – ident: e_1_2_8_45_1 doi: 10.1109/CVPR.2007.383157 – ident: e_1_2_8_58_1 doi: 10.1109/ISBI.2006.1625167 – ident: e_1_2_8_56_1 doi: 10.1088/0031‐9155/40/5/010 – ident: e_1_2_8_23_1 doi: 10.1088/0031‐9155/61/19/7092 – ident: e_1_2_8_27_1 doi: 10.1016/j.media.2015.08.001 – volume: 15 start-page: 1929 year: 2014 ident: e_1_2_8_54_1 article-title: Dropout: A simple way to prevent neural networks from overfitting publication-title: J. Mach. Learn. Res. – ident: e_1_2_8_60_1 – ident: e_1_2_8_39_1 doi: 10.1118/1.1800713 – ident: e_1_2_8_41_1 doi: 10.1117/12.2216327 – volume-title: Digital Mammography; IWDM 2000 year: 2001 ident: e_1_2_8_35_1 – ident: e_1_2_8_37_1 doi: 10.1118/1.2237543 – ident: e_1_2_8_9_1 doi: 10.1118/1.4770279 – ident: e_1_2_8_25_1 doi: 10.1118/1.4944498 – ident: e_1_2_8_26_1 doi: 10.1080/21681163.2015.1124249 – ident: e_1_2_8_55_1 doi: 10.1118/1.1758791 – ident: e_1_2_8_19_1 doi: 10.1088/0031‐9155/59/23/7457 – ident: e_1_2_8_43_1 doi: 10.1118/1.3284974 – ident: e_1_2_8_44_1 doi: 10.1118/1.598389 – ident: e_1_2_8_34_1 doi: 10.1117/12.878214 – start-page: 3320 volume-title: Advances in Neural Information Processing Systems year: 2014 ident: e_1_2_8_3_1 |
| SSID | ssj0006350 |
| Score | 2.6081786 |
| Snippet | Purpose:
Develop a computer-aided detection (CAD) system for masses in digital breast tomosynthesis (DBT) volume using a deep convolutional neural network... Purpose: Develop a computer‐aided detection (CAD) system for masses in digital breast tomosynthesis (DBT) volume using a deep convolutional neural network... Develop a computer-aided detection (CAD) system for masses in digital breast tomosynthesis (DBT) volume using a deep convolutional neural network (DCNN) with... PURPOSEDevelop a computer-aided detection (CAD) system for masses in digital breast tomosynthesis (DBT) volume using a deep convolutional neural network (DCNN)... |
| SourceID | pubmedcentral proquest pubmed crossref wiley scitation |
| SourceType | Open Access Repository Aggregation Database Index Database Enrichment Source Publisher |
| StartPage | 6654 |
| SubjectTerms | Analysis of texture Artificial neural networks Biological material, e.g. blood, urine; Haemocytometers Breast Neoplasms - diagnostic imaging cancer computer‐aided detection convolutional neural network deep‐learning Diagnosis, Computer-Assisted digital breast tomosynthesis Digital computing or data processing equipment or methods, specially adapted for specific applications Digital tomosynthesis mammography feature extraction Film mammography image classification Image data processing or generation, in general Image detection systems Image Processing, Computer-Assisted - methods image segmentation image texture In which a programme is changed according to experience gained by the computer itself during a complete run; Learning machines Inference methods or devices Learning learning (artificial intelligence) Machine Learning Mammography mass medical image processing neural nets Neural Networks (Computer) QUANTITATIVE IMAGING AND IMAGE PROCESSING radiology Segmentation Testing procedures transfer learning tumours |
| Title | Mass detection in digital breast tomosynthesis: Deep convolutional neural network with transfer learning from mammography |
| URI | http://dx.doi.org/10.1118/1.4967345 https://onlinelibrary.wiley.com/doi/abs/10.1118%2F1.4967345 https://www.ncbi.nlm.nih.gov/pubmed/27908154 https://www.proquest.com/docview/1845822987 https://pubmed.ncbi.nlm.nih.gov/PMC5135717 |
| Volume | 43 |
| WOSCitedRecordID | wos000390237200043&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D |
| hasFullText | 1 |
| inHoldings | 1 |
| isFullTextHit | |
| isPrint | |
| journalDatabaseRights | – providerCode: PRVWIB databaseName: Wiley Online Library - Journals customDbUrl: eissn: 2473-4209 dateEnd: 99991231 omitProxy: false ssIdentifier: ssj0006350 issn: 0094-2405 databaseCode: DRFUL dateStart: 19970101 isFulltext: true titleUrlDefault: https://onlinelibrary.wiley.com providerName: Wiley-Blackwell |
| link | http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwpV1Zb9RADLZ6QOGFo1zhqIZDiJdAkp0TnhBlxUO3qhCV9i1K5iiR2mS1SSv13zOeZINWLQiJpzzEiZKMHX8e258B3hihNPZHxpyLLKaO2biU1MYFp0j_lBhuA2X-gTg8lPO5OtqAT6temJ4fYtxwQ8sI_2s08KIcppCkWLievqeKiwllm7CNTVU-8tre_z49Phh_xN6X9h0oimISgQ3EQv7yD-PF6-7oCsa8Wip5y3umPkm-DmeDP5re_a83uQd3BhhKPvd6cx82bL0LO7Mh0b4LN0NlqG4fwOXMw2tibBdqtmpS1cRUJzhqhJRY0N6Rrjlr2svaI8m2aj-SfWsXBIvZB6X2gkiaGQ6h5Jzg3i_pAmK2SzLMrTgh2OlCzgpvGD2L9kM4nn798eVbPMxriDX1QC42xgdPSqTKOqGldjrNslRnpeCFEDzRynk0qNmk5EwUzgMfarRxE8mtVNJ5F_oItuqmtk-ASEwfCpmyghdUWVs4lwhhTJLoxFCTRPButWz5aj1wpsZp3gc1Mk_z4aNG8GoUXfQMHtcJvVytfe7tC5MmRW2b8zb3ETBDUnwpInjc68J4m0woj6gYjUCsackogNzd62fq6mfg8GbphPlIOoLXoz797emukbpolr8l8oVxEbwNWvbn--SzIzw8_VfBZ3Db40PeV-88h61ueW5fwA190VXtcg82xVzuDUb3C0vdLcQ |
| linkProvider | Wiley-Blackwell |
| linkToHtml | http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwpV1Zb9QwEB6VLbS8cJQC4TSHEC-BHD4RL4iyKmJ3VaFW6luU9VEi0WS1SSv13-MjG7RqQUg8-SFjK4lnPJ89428AXismpLsfGVPKshgbouM5xzouKXb0T4mi2lPmT9hsxo-PxcEGfFzdhQn8EMOBm7MMv147A3cH0r2Vu8z19B0WlOWYXINNTHPGR7C59318NBlWYutMwxUUgV0UgfTMQrb7-6Hzuj-6BDIv50puW9cUouTreNY7pPHt__uUO3CrB6LoU9Ccu7Ch6x3Ymvah9h244XNDZXsPLqYWYCOlO5-1VaOqRqo6ccVG0NyltHeoa06b9qK2WLKt2g9oT-sFcunsvVpbQUeb6RufdI7c6S_qPGbWS9RXrjhB7q4LOi2taQQe7V04Gn85_Lwf9xUbYoktlIuVstsnwVKhDZNcGplmWSqzOaMlYzSRwlg8KEk-p4SVxkIfrKQyOaeaC26sE70Po7qp9UNA3AUQGU9JSUsstC6NSRhTKklkorBKIni7mrdiNSGuqsbPImxreJEW_U-N4OUguggcHlcJvVhNfmEtzIVNylo3Z21h98DE0eJzFsGDoAzDMBkTFlMRHAFbU5NBwLF3rz-pqx-exZukObF76QheDQr1t7e7Quq8Wf6WKBbKRPDGq9mfxymmB6559K-Cz2F7_3A6KSZfZ98ew02LFmnI5XkCo255pp_CdXneVe3yWW97vwAuGDDM |
| linkToPdf | http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwpV1Lb9QwEB6VLRQuPFoK4WkeQr0EkqyfiAtiWYHYXa0QlXqzsn6USG2y2qSV-u-xnWzQqgUhcfIhYyuJZzyfPeNvAF5rJpS_HxlTyrIYW2LiBccmzin29E-JpiZQ5k_YbMaPjsR8Cz6s78K0_BD9gZu3jLBeewM3S207K_eZ6-lbLCgbYnINtjERBA9ge_R9fDjpV2LnTNsrKAL7KALpmIVc93d9501_dAlkXs6VvOlcUxsl38SzwSGN7_zfp9yF2x0QRR9bzbkHW6bchZ1pF2rfhRshN1TVe3AxdQAbadOErK0SFSXSxbEvNoIWPqW9QU11WtUXpcOSdVG_RyNjlsins3dq7QQ9bWZoQtI58qe_qAmY2axQV7niGPm7Lug0d6bR8mjfh8Px5x-fvsRdxYZYYQflYq3d9kmwVBjLFFdWpVmWqmzBaM4YTZSwDg8qMlxQwnLroA_WStshp4YLbp0T3YdBWZXmISDuA4iMpySnORbG5NYmjGmdJCrRWCcRHKznTa4nxFfVOJHttobLVHY_NYKXveiy5fC4SujFevKlszAfNslLU53V0u2BiafF5yyCB60y9MNkTDhMRXAEbENNegHP3r35pCx-BhZvkg6J20tH8KpXqL-93RVS59Xqt4R0GhXBm6Bmfx5HTue-efSvgs9hZz4ay8nX2bfHcMuBRdqm8jyBQbM6M0_hujpvinr1rDO9X3MeMEc |
| openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Mass+detection+in+digital+breast+tomosynthesis%3A+Deep+convolutional+neural+network+with+transfer+learning+from+mammography&rft.jtitle=Medical+physics+%28Lancaster%29&rft.au=Samala%2C+Ravi+K&rft.au=Chan%2C+Heang-Ping&rft.au=Hadjiiski%2C+Lubomir&rft.au=Helvie%2C+Mark+A&rft.date=2016-12-01&rft.eissn=2473-4209&rft.volume=43&rft.issue=12&rft.spage=6654&rft.epage=6654&rft_id=info:doi/10.1118%2F1.4967345&rft.externalDBID=NO_FULL_TEXT |
| thumbnail_l | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0094-2405&client=summon |
| thumbnail_m | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0094-2405&client=summon |
| thumbnail_s | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0094-2405&client=summon |